V/ AS
8332-
Volume 101
Number 1
Spring 2015
Journal of the
WASHINGTON
ACADEMY OF SCIENCES
MCZ
LIBRARY
NOV 3 0 2015
HARVARD
UNIVERSITY
Board of Discipline Editors
Editor’s Comments S. Rood iii
Intellectual Washington Today S. Umpleby 1
Does Speed Matter? The Employment Impact of Increasing Access to
Fiber Internet P. Lapointe 9
Benjamin Banneker and Celestial Navigation: Just How Did They Know
Where They Were, Then? S. Howard 29
Washington Academy of Sciences Awards Program 2015 43
Addendum to Washington Academy of Sciences 2014 Membership Directory 53
In Memoriam: Burton G. Hurdle (1918-2015) 57
Membership Application 59
Instructions to Authors 60
Affiliated Institutions 61
Affiliated Societies and Delegates 62
ISSN 0043-0439
Issued Quarterly at Washington DC
Washington Academy of Sciences
Founded in 1898
Board of Managers
Elected Officers:
President
Mina Izadjoo
President Elect
Mike Coble
Treasurer
Ronald Hietala
Secretary
John Kaufhold
Vice President, Administration
Nick Tran
Vice President, Membership
Sue Cross
Vice President, Junior Academy
Vice President, Affiliated Societies
Gene Williams
Members at Large
Paul Arveson
Michael P. Cohen
Frank Haig, S.J.
Neal Schmeidler
Mary Snieckus
The Journal of the Washington Academy of
Sciences
The Journals the official organ of the Academy.
It publishes articles on science policy, the history
of science, critical reviews, original science
research, proceedings of scholarly meetings of
its Affiliated Societies, and other items of interest
to its members. It is published quarterly. The last
issue of the year contains a directory of the
current membership of the Academy.
Subscription Rates
Members, fellows, and life members in good
standing receive the Journal free of charge.
Subscriptions are available on a calendar year
basis, payable in advance. Payment must be
made in U.S. currency at the following rates.
U.S. and Canada $30.00
Other Countries $35.00
Single Copies (when available) $15.00
Claims for Missing Issues
Claims must be received within 65 days of
mailing. Claims will not be allowed if non-
delivery was the result of failure to notify the
Academy of a change of address.
Past President
Terrell Erickson
Affiliated Society Delegates
Shown on back cover
Editor of the Journal
Sally A. Rood
Notification of Change of Address
Address changes should be sent promptly to the
Academy office. Notification should contain both
old and new addresses and zip codes.
POSTMASTER:
Send address changes to Washington Academy
of Sciences, Room 113, 1200 New York Ave.
NW, Washington, DC 20005
Academy Office
Washington Academy of Sciences
Room 113
1 200 New York Ave. NW
Washington, DC 20005
Phone: (202) 326-8975
Journal of the Washington Academy of Sciences
(ISSN 0043-0439)
Published by the Washington Academy of
Sciences (202) 326-8975
Email: iournal@washacadsci.ora
Website: www.washacadsci.ora
1
Journal of the
1200 New York Ave.
Suite 113
Washington DC
20005
www.washacadsci. org
MCZ
LIBRARY
WASHINGTON NOV 3 0 2015
ACADEMY OF SCIENCES HARVARD
UNIVERSITY
Volume 101 Number 1 Spring 2015
Contents
Board of Discipline Editors ii
Editor’s Comments S. Rood iii
Intellectual Washington Today S. Umpleby 1
Does Speed Matter? The Employment Impact of Increasing Access to
Fiber Internet P. Lapointe 9
Benjamin Banneker and Celestial Navigation: Just How Did They Know
Where They Were, Then? S. Howard 29
Washington Academy of Sciences Awards Program 2015 43
Addendum to Washington Academy of Sciences 2014 Membership Directory ....53
In Memoriam: Burton G. Hurdle (1918 - 2015) 57
Membership Application 59
Instructions to Authors 60
Affiliated Institutions 61
Affiliated Societies and Delegates 62
ISSN 0043-0439 Issued Quarterly at Washington DC
Spring 2015
ii
Journal of the Washington Academy of Sciences
Editor Sally A. Rood, PhD sally.rood2@gmail.com
Board of Discipline Editors
The Journal of the Washington Academy of Sciences has a 12-
member Board of Discipline Editors representing many scientific and
technical fields. The members of the Board of Discipline Editors are
affiliated with a variety of scientific institutions in the Washington area
and beyond — government agencies such as the National Institute of
Standards and Technology (NIST); universities such as George Mason
University (GMU); and professional associations such as the Institute of
Electrical and Electronics Engineers (IEEE).
Anthropology
Astronomy
Biology/Biophysics
Botany
Chemistry
Environmental Natural
Sciences
Health
History of Medicine
Operations Research
Physics
Science Education
Systems Science
Emanuela Appetiti eappetiti@hotmail.com
Sethanne Howard sethanneh@msn.com
Eugenie Mielczarek mielczar@, physics. gmu.edu
Mark Holland maholland@salisbury.edu
Deana Jaber diaber@marvmount.edu
Terrell Erickson terrell.ericksonl@wdc.nsda.gov
Robin Stombler rstombler@auburnstrat.com
Alain Touwaide atouwaide@hotmail.com
Michael Katehakis mnk@rci.rutgers.edu
Katharine Gebbie katharine.gebbie@nist.gov
Jim Egenrieder iim@deepwater.org
Elizabeth Corona elizabethcorona@gmail.com
Washington Academy of Sciences
Ill
Editor’s Comments
In this issue of the Journal of the Washington Academy of Sciences
we are celebrating the Washington, D.C., region and its science and
technology presence!
Back in 1985, Amitai Etzioni’s Washington Post editorial, “The
World-Class University that Our City Has Become,” was his personal
statement as a new resident of the Washington, D.C., area in the mid-
1980s. It provided an interesting view of the city’s aspirations in science
and technology and policy circles at that time. Stuart Umpleby
rediscovered this editorial and provides an updated perspective in
“Intellectual Washington Today.” While Etzioni’s emphasis was on the
policy community — he called it “Washington Metropolitan University”
or W.M.U. — Umpleby’ s emphasis is on the more recent growth of
information-related activities in the Washington, D.C. area. Regardless,
these dual perspectives highlight the important role of the science and
technology community and academic and policy institutions in the affairs
of the Washington, D.C., metropolitan region.
In line with this celebration of the Washington area’s science
presence, it is fitting that this issue documents the Academy’s annual
Awards Program and ceremony. Sethanne Howard presented the keynote
at the banquet — about the scientist, Benjamin Banneker, who lived in the
Baltimore area from 1731 to 1806. The geographical boundary for
Washington, D.C., was surveyed back in the late 1700s using the eclipses
of the Galilean satellites to determine longitude. As part of the survey
team, Banneker timed the eclipses of the Galilean satellites by Jupiter and
he kept the survey clocks running at the right time. Thus, the title of the
presentation at the awards ceremony was “Benjamin Banneker and
Celestial Navigation: Just How Did They Know Where They Were,
Then?”. Banneker’s story was quite interesting at the banquet ... and is
now equally interesting in this issue of the Journal !
As an introduction to the Academy’s 2015 Awards Program, we
provide a “backgrounder” on the Awards Program. It includes an early
history of the program, some traditions, and a primer on the nominations
process.
Congratulations to these distinguished scientists and educators in
Washington, D.C., area scientific institutions whom the Academy honored
with awards in their fields this year: Ronald Colie, Ram D. Sriram,
Marcus Cicerone, Paul Peterson, Robert Gover, Gregory Strouse, and
Spring 2015
IV
MaryBeth Petrasek. Details on each of their awards are presented along
with photos from the ceremony and banquet.
A quantitative study of the policy implications of broadband
improvements across the country is presented in the paper by Paul
Lapointe entitled “Does Speed Matter? The Employment Impact of
Increasing Access to Fiber Internet.” The study finds a positive association
between increasing access to fiber cable and increases in employment and
the number of firms at the county level which, in turn, offers evidence that
promoting access to fiber internet is a viable approach to economic
development.
This Journal issue also includes an Addendum to the Academy’s
2014 Membership Directory which appeared in the Winter 2014 issue of
the Journal of the Washington Academy of Sciences. The names of twenty
new members who were inadvertently omitted from the 2014 Membership
Directory are printed in this issue instead of waiting for the next Directory.
My sincere apologies for their omission from our annual Membership
Directory this past year.
In addition, in this Journal issue we honor the life of Burton
Hurdle, a long-time member of the Washington science community who
passed away this Spring.
Finally, this Journal issue marks my last issue as editor. I’ve edited
the Journal for three years, and at this time I am handing over the
editorship to Sethanne Howard. Please send Sethanne your manuscripts
and other input going forward. Eve enjoyed working with all the authors,
reviewers, proofreaders, and members of the Board of Discipline Editors
who have contributed their time so that we can maintain high standards for
the Journal. I’ve been blessed by the large number of talented people
interested in supporting this unique peer-reviewed interdisciplinary
Journal. It’s been an honor working with all of you ... too many to
mention individually over that period of time ... please know that I
appreciate and thank each of you from my heart.
Sally A. Rood, PhD, Outgoing Editor
Journal of the Washington Academy of Sciences
Washington Academy of Sciences
1
Guest Editorial
Intellectual Washington Today
Stuart A. Umpleby
The George Washington University, Washington, D.C.
Abstract
In a Washington Post editorial thirty years
ago, Amitai Etzioni described how
Washington, D.C. was becoming an
intellectual city. Previously, Washington
was viewed as the home of the national
government, journalism, lawyers, and
lobbyists but not as an academic or
intellectual city. However, Etzioni claimed
that Washington had become a policy and
scientific powerhouse as a result not only
of its growing and improving universities
and their research institutes, but also
because of its federal agencies, think tanks, and policy research
organizations. This article reviews the points made by Etzioni and
examines the situation today.
Introduction
Washington, D.C., is a city with many ironic descriptions. It is often
described as the Northern-most Southern city. John F. Kennedy said it was
a city of Northern charm and Southern efficiency. It has been called a city
full of former student body presidents, and a city consisting of residents
who come from somewhere else. Currently Washington may be known as
a city of politicians, interns, diplomats, and bloggers. It is not often
thought of as a scientific city or an intellectual city. However, Washington
has been growing and changing. As the nation becomes increasingly a
post-industrial society, Washington is becoming a leader in new types of
organizations and new kinds of jobs.
A Description of Washington 30 Years Ago
To explain the purpose of an editorial he contributed to the
Washington Post in the Spring of 1985, Amitai Etzioni said that people
sometimes asked him why he had moved from Columbia University to
Washington, D.C., which previously had not had a reputation as a source
Spring 2015
2
of innovative ideas. He wrote that the Washington Metropolitan
University — the combination of universities, policy research institutes,
and government agencies — “easily matches the intellectual vigor of
contemporary London,” and that it had “almost as many little magazines
(where intellectuals float new ideas) and writers-in-residence as the Left
Bank of Paris.”
Etzioni pointed out that several new research organizations had
been added to the D.C. area prior to 1985: the Roosevelt Center, the
Center for National Policy, and the Cato Institute.
He also noted that the National Institutes of Health (NIH) did more
research in biology and related disciplines than was conducted at Harvard,
Yale, Princeton, Columbia and Brown combined.
Major research centers in economics could be found in the World
Bank, the Federal Reserve Board, and the Congressional Budget Office.
The natural sciences were strong in the Carnegie Institution of
Washington, the Smithsonian Institution, the National Institute of
Standards and Technology in Gaithersburg, Maryland, and the Department
of Defense (DOD).
Etzioni made a distinction between academics who were deep
scholars of narrow topics and intellectuals who took a broader perspective
on the direction of American society and trends in literature and the arts.
He claimed that many intellectuals had moved to D.C. because they found
the academic abundance congenial.
Etzioni also noted that academics and intellectuals communicated
with each other not only in seminars, but also in magazines that stimulated
new ideas. As just a few of these published in D.C., he listed:
• The Wilson Quarterly ,
• the American Enterprise Institute’s Public Opinion ,
• Regulation , which reports on the effects of government
intervention,
• The Cato Journal, and
• Foreign Policy magazine, then a new competitor to Foreign
Affairs, published in New York.
Etzioni further noted that Science magazine was the nation’s leading
journal of science, and that the National Academy of Sciences published
Washington Academy of Sciences
3
Issues if 7 Science and Technology > (both products of D.C.) Finally, he
noted that Washington provides numerous television news and discussion
programs to the nation.
As an academic and intellectual city, how has Washington
progressed since 1985?
Many Universities in Washington
There has been continued growth and improvement in universities,
particularly the growth of George Mason University since it became
independent in 1972.
The familiar Washington, D.C., universities — American
University, Catholic University, Georgetown University, George
Washington University, Howard University, Johns Hopkins University’s
School of Advanced International Studies, the University of Maryland,
Marymount University, and the University of the District of Columbia —
are all prospering.
Several well-established universities, for example George
Washington University and the University of Maryland, now have
buildings in several parts of the city. These locations provide classes more
conveniently to students but also conduct research, as does George
Washington University’s Virginia Science and Technology Campus in
close-by Ashbum, Virginia.
Universities based in other cities also have a presence in the
Washington area. For example, Cornell University, New York University,
Syracuse University, Pepperdine University, and Virginia Tech are all
here.
Clearly universities find it desirable to have a connection to
Washington, D.C.
The Growth of Information-Based Activities
The information intensive activities of the federal government have
also expanded greatly since 30 years ago. A few examples of such
activities in the Washington area are the following:
• The National Security Agency at Fort Meade, Maryland, has
become the center of a “cyber valley” in the Baltimore -
Washington corridor. [Schiff, 2013]
Spring 2015
4
• The Dulles access toll road in northern Virginia contains the
expanded “beltway” contracting firms and information services
firms such as AOL.
• The Route 270 corridor in Maryland just north of D.C. continues to
be the home for biological research, with key institutions being
NIH, the Walter Reed Army Medical Center, and Bethesda Naval
Medical Center.
• Research programs, administered at NASA Headquarters and the
Goddard Space Flight Center in nearby Greenbelt, Maryland, have
made fundamental contributions to improving weather forecasting,
to earth science, and to our understanding of climate change.
NASA’s Hubble Space Telescope has dramatically advanced our
understanding of the cosmos.
• The number of patents issued by the U.S. Patent and Trademark
Office (USPTO), headquartered in Alexandria, Virginia, in the past
twenty years has more than tripled, from 113,268 in 1994 to
329,613 in 2014. [USPTO, 2015] The USPTO now has not only a
new building but a new campus in Alexandria, just south of D.C.
Washington is definitely a leader in applications of information
technology. The Internet, an outgrowth of an earlier DOD research
project, has transformed business, government and personal
communication in recent years. The D.C. area’s knowledge workers now
spend hours each day in “cyberspace” and the contents of filing cabinets
are now “in the cloud” with both positive and negative consequences.
Cybersecurity is a leading domestic and international concern and
“identity theft” is a new worry for private citizens.
The Washington Post is now owned by Amazon.com. Many
newspapers have gone out of business. There are now numerous blogs
written by former journalists.
Improving Management in Government and Business
In the years since Etzioni wrote his article, there have been many
changes in the federal government which have transformed both the
practice of government in Washington and also influenced the
management of corporations and state and local governments.
In 1987, Congress created the Malcolm Baldrige National Quality
Improvement Program aimed at improving the productivity of U.S. firms,
Washington Academy of Sciences
5
which in the 1970s were having difficulty competing with Japanese
manufacturers. The Baldrige National Quality Award was expanded to
include education and health care organizations in 1 999, and a government
and non-profit category was added in 2007.
As an example of Washington’s growing influence, a 1991 General
Accounting Office report showed how the Baldrige Program companies
increased their market share an annual average of 13.7 percent. [Garvin,
1991] Such a high growth rate meant that companies using quality
improvement methods in just a few years bought or replaced companies
not using these methods. A more recent study said that participating
companies had an 820:1 ratio of benefits for the U.S. economy to program
costs. [Link and Scott, 2012] To arrive at this ratio, they compared the
benefits received by the 273 Malcolm Baldrige National Quality Award
applicants from 2007 to 2010 with the cost of operating the Baldrige
Program. The 820-to-l ratio represents only the benefits for the surveyed
applicants, but it represents all of the Baldrige Program’s costs. Link and
Scott note that the benefit-to-cost ratio would be much higher if program
costs were compared with benefits for the entire U.S. economy.
Quality Improvement Methods were taken seriously by President
Bill Clinton who appointed Vice President A1 Gore to head the National
Performance Review in 1993. This initiative had the goal of dramatically
improving the performance of the federal government through a
combination of process improvement methods and increased contracting
as an alternative to larger government agencies.
In March 1998, the National Performance Review pointed to a
number of important achievements, later presented in Kamensky [1999]:
• The size of the federal civilian workforce was cut by 351,000 —
the smallest since President Kennedy held office and, as a
percentage of the national workforce, the smallest since 1931.
• Action was recommended on about 1,500 issues in 1993 and 1995.
Agencies completed about 58 percent. Of the original
recommendations, 66 percent were reported as completed. For
those requiring Presidential or Congressional action, President
Clinton signed 46 directives and Congress passed and the President
signed over 85 laws.
• About $177 billion in savings were recommended over a 5-year
period. Agencies locked into place about $137 billion. In addition,
as of March 1998, the process improvement award winners
Spring 2015
6
estimated savings or cost avoidances of about $3 1 billion because
of their actions.
• Agencies eliminated about 640,000 pages of internal rules, about
16,000 pages of Federal Regulations, and rewrote 31,000
additional pages into plain language.
• Agencies sponsored 850 labor-management partnerships. A 1998
survey of employees showed those in organizations that actively
promoted reinvention were twice as satisfied with their jobs.
• Over 570 federal organizations had committed to more than 4,000
customer service standards.
Kamensky [1999] also reported that public trust in the federal government
was increasing after a 30-year decline. While it was not clear whether this
improvement was directly linked to the work of the National Performance
Review, the Review made an important contribution.
The Use of Information in Policy-Making
Who analyzes information and writes reports in the D.C. area has
also been changing. Since Etzioni wrote his article, the Office of
Technology Assessment was closed and the number of Congressional staff
was significantly reduced during the time that Newt Gingrich was Speaker
of the U.S. House of Representatives. As a result, it can be said that the
task of providing background information for legislation has been taken up
by lobbying firms on K Street which often draft new legislation, a task
previously performed by Congressional staff members. [Benen, 2011]
Also, political activity has moved from public demonstrations on
the mall to campaign contributions and lobbying behind closed doors. It is
harder now, in 2015, to know what changes in laws are occurring. Hedrick
Smith [2012] in his recent book notes that when he was head of the
Washington bureau of The New York Times, he did not realize that a series
of laws and court decisions were fundamentally changing taxes and
entitlement programs beginning in the late 1970s. Over time, these
changes have led to a dramatic increase in inequality in the United States,
which has affected all U.S. citizens.
Conclusion
In his article 30 years ago, Amitai Etzioni focused primarily on the
many policy research organizations in Washington. During his years as a
professor at the George Washington University, Etzioni himself has made
Washington Academy of Sciences
7
notable contributions to policy research and discussions. He created the
Society for the Advancement of Socio-Economics and an academic
journal, the Journal for the Advancement of Socio-Economics. He founded
and leads the Communitarian Network, a non-profit, non-partisan
organization dedicated to supporting the moral, social and political
foundations of society. He is currently Director of the Institute for
Communitarian Policy Studies at George Washington University.
Of course not all of the information-related activities in the
Washington area involve transformative policy analyses. Much of the
work — for example, at the Patent Office and the National Security
Agency — requires careful attention to detail. In the past 30 years, the
number of information-related jobs in the Washington, D.C., area has
increased dramatically.
However, large organizations that conduct these information
processing activities create a demand for educated workers and, just as
importantly, for additional innovations in handling information-related
tasks. For this reason, several local universities have recently started
degree programs in big data, data analytics, and cyber security.
The “post-industrial” society which has exploded in the D.C. area
in the past 30 years has also been growing globally. Around the world,
new universities are being established and are improving. In any event,
Washington, D.C., is well-positioned to be a leading city in this post-
industrial era.
Overall, the city-wide university that Etzioni described 30 years
ago is a key player in defining and creating the nation and the world in the
21st Century.
References
Benen, Steve. 2011. Lobbyists Go Back to Writing Laws, Washington
Monthly , March 18.
Etzioni, Amitai. 1985. The World-Class University That Our City Has
Become. The Washington Post. April 28.
Garvin, David. 1991. How the Baldrige Award Really Works. Harvard
Business Review. November-December, pp. 80-94.
Spring 2015
8
Kamensky, John. 1999. National Partnership for Reinventing Government
(formerly the National Performance Review): A Brief History.
Washington, D.C. 20006.
http://govinfo.hbrarv.unt.edu/npr/whoweare/history2.html
Link, Albert N. and Scott, John T. 2012. On the Social Value of Quality:
An Economic Evaluation of the Baldrige Performance Excellence
Program. Science & Public Policy , 39, 5: 680-689.
Schiff, Philip. 2013. Commentary: How to build a regional ‘Cyber Valley’
in Capital Business, The Washington Post Co.
Smith, Hedrick. 2012. Who Stole the American Dream? Random House,
2012.
U.S. Patent and Trademark Office. 2015. Performance and Accountability
Report, http://www.uspto.gov/about-us/perfonuance-and-
planning/uspto-annual-reports.
Bio
Stuart A. Umpleby is Professor Emeritus in the School of
Business at The George Washington University in Washington, D.C. He
may be contacted at umplebv@gwu.edu. www.gwu.edu/-umplebv.
Washington Academy of Sciences
9
Does Speed Matter? The Employment Impact of
Increasing Access to Fiber Internet
Paul Lapointe
Georgetown University, Washington, D.C.
Abstract
As internet technology continues to improve at a rapid pace, there is
constant debate over the relative value of various internet connection
technologies. In recent years, policymakers have debated over several
important questions regarding broadband. What speed qualifies as high-
speed broadband? How much public funding should be spent increasing
access to broadband? And, what regulations to impose on internet
providers? While the potential and proven benefits of high-speed
internet are diverse, the economic impacts are often at the forefront of
policy discussions. To date, most research into the economic impact of
internet has focused on increasing access to and adoption of broadband
internet in general, without emphasizing the speed of the broadband
connections. This paper utilizes new data available as a result of the
American Recovery and Reinvestment Act to examine the relationship
between employment growth and access to fiber internet, currently seen
as the gold standard of internet connections in terms of speed and
reliability. Using data from the National Broadband Map, this study
finds a positive association, within the United States, between
increasing access to fiber and increases in employment and number of
firms at the county level. This positive relationship provides evidence
to policymakers that promoting access to fiber internet is a viable
economic development approach.
Introduction
Although there is a strong consensus that high-speed internet is
related to economic growth, many questions remain about what speed is
optimal. As the internet becomes ubiquitous in the United States, attention
has shifted from expanding access to the internet towards improving the
connections that Americans have access to. Table 1 shows the percent of
U.S. households that have access to different types of internet
technologies. Almost all households have access to some form of internet
connection, whether it is a fixed line connection, wireless internet, or
satellite. Additionally, 95 percent of households have access to fixed line
internet, including 87 percent that have access to a cable internet
connection. The opportunity that remains is in expanding access to state of
the art technologies such as optical fiber, where access is expanding in
Spring 2015
10
recent years, but still remains out of reach for most American households,
as only one in four households has access to it.
Table 1. Percent of U.S. Households with Access to Internet Technologies in June
201 1 and December 2013
Over the past two decades, the policy focus has been on
increasing broadband access and adoption. In the aggregate, these efforts
have largely been successful. Broadband access (using the Organisation
for Economic Co-operation and Development (OECD) definition of 256
Kbit/sec) in the United States has increased from 4.4 percent of
households having access in 2000 to 19.9 percent of households in 2003
and 68.2 percent in 2010 (OECD, 2014). Now that broadband access is
wider, many policymakers have shifted away from increasing access
towards increasing speed. The demand for high speed is clear; when
Google announced plans to pilot its Google Fiber networks, more than
1,100 communities across the country applied (Kelly, 2010). Absent
private investment, some municipalities have dedicated vast tax payer
resources to construct fiber networks of their own. Clearly, effort is being
put into improving internet connections, yet there is little empirical
evidence as to whether these ultra-high-speed networks promote growth
beyond the benefits of more common speeds. The purpose of this paper is
to examine the economic impact of fiber internet availability in the United
States.
Now, thanks to recent enhancements to the Federal
Communication Commission’s (FCC) data collection strategies,
researchers have access to data which allows the examination of the
impact of fiber networks for the first time. By evaluating the economic
impact of fiber internet, information can be provided to policymakers to
help guide them in determining the amount of resources to invest in the
technology.
Washington Academy of Sciences
11
Literature Review
There is a growing body of literature on the economic impact of
high speed internet both in the United States and around the world; the
general consensus is that high speed internet leads to economic growth
(Qiang, 2009; Van Gaasbeck, 2008; Whitacre et al 2013; Kolko, 2012).
The literature can be divided into research on differences in broadband
technology across countries and differences in broadband technology
within a single country. While this paper will focus on the effect of
broadband differences in the United States, it is important to examine the
literature in both areas to build a cohesive picture of the state of research
on economic effects of broadband.
International Literature
As a whole, the literature on country-level effects of broadband
technology shows that countries with higher levels of broadband
penetration have generally higher levels of GDP growth. Czemich et al.
(2009) used data from a panel of 25 OECD countries between 1 996 and
2007 to create a model — using pre-existing telephone and TV networks
to predict maximum broadband penetration rates — to examine economic
impact. They (2009) found a statistically significant positive relationship;
a 10 percent increase in broadband penetration raised GDP per capita by
0.9- 1.5 percent.
Similarly, Qiang and Rossotto (2009) used Information
Communications and Technologies Development (ICTD) and World Bank
data for 120 countries between 1980 and 2006 to understand how differing
broadband penetration rates are related to GDP per capita growth. They
estimated that a 10 percent increase in broadband adoption is associated
with a 1.21 percent increase in GDP per capita for developing countries
and a 1.38 percent increase for developed countries. However, Qiang and
Rossotto (2009) caution that causality is not abundantly clear; that is, there
could be a back and forth effect as increased wealth also increases the
demand for broadband services. Koutroumpis (2009) attempted to account
for the fact that broadband can both influence and be influenced by
economic factors using a simultaneous equation model to identify the
macro impact of broadband in 15 European Union countries between 2003
and 2006. He separated the increased demand for broadband caused by
increased wealth from the economic growth caused by increased
broadband usage with models that predict the supply and demand for
broadband growth. After separating out these influences, there was still a
Spring 2015
12
significant, positive relationship between broadband penetration and GDP
per capita.
National Literature
Research within the United States has been building over time,
with researchers using a variety of datasets and approaches to building
models. While the approaches vary, there is a consensus that
improvements in broadband technology are related to higher levels of
employment, although findings are mixed on other economic indicators
such as number of firms and income.
The early literature in the United States focused on building cross-
sectional panel models that take advantage of varying levels of
technological development across regions of the country. Lehr et al.
(2005) used data from the FCC form 477 and the Population Censuses and
Establishments Surveys to investigate the effect of broadband presence (as
a binary measure) on economic indicators such as employment, wages,
and industry mix. Their model, which used data from 1998-2002, showed
that in zip codes with mass-market broadband availability there was higher
employment, more firms overall, and more firms in the IT sector. The
broadband speed studied was 200 kilobits per second, speeds that would
now be considered slow. Moreover, the Lehr et al. (2005) study showed
the tradeoffs associated when choosing to study broadband at the state or
community level in the United States. Crandall et al. (2007) built on this
model with data from 2003-2005 to examine state level GDP growth
associated with increased broadband penetration. While they found that
higher levels of broadband penetration were associated with higher levels
of GDP growth, the results were not statistically significant, which
reinforces the notion that state-level data are too broad to study broadband
in America. While several dependent variables of interest, such as GDP,
are not available at smaller geographical units than the state, there is
generally not sufficient variation between states in broadband availability
to draw meaningful conclusions.
While much of the research in the United States uses FCC data,
two studies in 2007 corroborate the larger national studies using different
data sources. Van Gaasbeck (2008) used cross-sectional panel household
survey data from Scarborough Research to examine the potential
employment effects of expanding broadband adoption in California. They
found that increased broadband adoption was associated with higher
employment but fewer establishments. Similarly, Shideler et al. (2007)
Washington Academy of Sciences
13
looked at county-level effects for a single state, Kentucky. They focused
on increased broadband availability, instead of broadband adoption. Using
infrastructure data from providers collected through ConnectKentucky,
they examined county-level employment growth and sector employment
growth relative to broadband availability, controlling for past growth,
education, unemployment, and road density. They found a positive,
statistically significant relationship between broadband availability and
total employment. While the limited scope of these studies restricts the
applicability to broader national policies, they help to validate the general
association between broadband and employment.
In a qualitative analysis, Ezell et al. (2009) made the case for
facilitating the development of internet with speeds of at least 20 Mbit/sec
downloading and preferably 50 Mbit/sec or greater. While most policy
efforts focus on increasing broadband adoption and availability, the
authors encourage policymakers to consider efforts to increase speeds as
well. They count fiber to the home, fiber to the node, and DOCSIS 3.0
cable as the most desirable fixed-line broadband delivery methods and 4G
as the most desirable wireless delivery method. They point out that
countries such as Japan, Singapore, South Korea, and Sweden are far
ahead of the United States in terms of high speed internet, giving them an
advantage in developing innovative web-based applications. In order for
the United States to remain the global leader in internet based innovation,
they contend that there needs to be a greater focus on increasing
broadband speed.
There has been a current focus on the impact of broadband
expansion in rural communities in particular. While high speed internet
has become standard in most urban and suburban communities, lower
population density makes it much more costly for providers to expand into
rural areas. Therefore, many policy initiatives have focused on how the
government can play a role in expanding access in rural communities.
Stenberg et al. (2009) match rural counties that had broadband by 2000
with those that did not based on a variety of characteristics in order to test
a causal relationship between broadband and economic growth in rural
counties. They aggregated FCC Form 477 data to measure broadband
availability and found faster employment growth in counties with more
availability. There is also evidence that counties which had early adoption
of broadband experienced relative income growth, but this faded over time
as broadband became more profuse. Whitacre et al. (2013) used data
newly available from the National Broadband Map combined with
adoption rates from FCC Form 477 to examine economic impacts of
Spring 20 1 5
14
broadband expansion into rural communities. They used three different
techniques to examine the relationship between broadband and economic
health. The collective results indicated a positive relationship between
rural economic indicators and broadband availability. They concluded that
adoption thresholds had more of an impact than availability thresholds
(Whitacre et at., 2013).
In regards to the debate over whether to use adoption or
availability as the key indicator of broadband penetration, Kolko (2012)
made the case for availability. He pointed out that adoption rates can be
influenced by economic growth more so than availability. Additionally,
increasing availability is a more feasible approach for policymakers than
increasing adoption rates. Using cross-sectional panel data from the FCC
between 1999 and 2006, Kolko built a model to identify the impact of
availability on local level employment and county-level labor market
outcomes. He found a statistically significant, positive relationship
between broadband expansion and local employment, but cautions that the
increased employment is accompanied by increased population growth,
resulting in no impact to employment rates.
In 2013, NC Broadband hosted a research roundtable to discuss the
state of research on the economic and community impact of broadband
expansion (Feser et al., 2013). The final report suggested that there is a
need for more research on specific broadband policies and investments at
the margin; including increases in broadband speeds and reliability and
use of new technology. This paper will attempt to fill some of that gap. It
benefits from the requirement in the National Broadband Plan that states
collect more detailed information on different technologies and speeds
available at local levels. Using this new dataset, it is now possible to start
evaluating whether or not incremental expansion of the presence of fiber
technology is associated with increased economic growth.
Study Hypothesis
The central hypothesis being tested in this study is that, within the
United States, increasing access to fiber internet connections is related to
increased levels of economic growth, as measured by employment levels,
number of firms, and income. Broadband, in general, can lead to economic
growth in several ways. By connecting individuals and companies across
the globe, the internet can make it easier for small and medium sized firms
to do business with suppliers and customers that they otherwise would not
have interactions with. Further, individuals are able to use the internet to
Washington Academy of Sciences
15
connect with employers and potential work remotely for companies
anywhere in the world, opening up more employment opportunities and
facilitating virtual talent mobility. Lastly, we would expect a short-term
increase in employment due to the fact that creating the connections
requires the hiring of employees to dig up cables, install new lines, and
provide on-going maintenance services. Because fiber internet provides a
faster, more reliable connection that allows the almost-instantaneous
transfer of large amounts of data, it is likely that these effects are enhanced
beyond what would be expected with more common speed levels.
Data
Much of the prior literature in the United States used FCC form
477 data to understand where broadband technology was available. While
this dataset provided a relatively complete picture, it did not offer insight
into different speeds within each geographical region. As part of the
American Recovery and Reinvestment Act of 2009, the National
Broadband Map was commissioned. The National Broadband Map
provided funding for each state to gather more detailed internet data. The
methodology used by each state to obtain these data differs slightly, but
there are set data fields that each state is required to provide. This semi-
annual data release is what allows the examination being conducted in this
study. The data are made available in several formats, such as the analyze
tables that aggregate internet statistics by region with accompanying
descriptive data that can be used in modeling efforts. The first analyze
table to be released was in 201 1 and it has been released every six months
subsequently.
Dependent variables for this model will come from the Quarterly
Census of Employment and Wages (QCEW) survey conducted by the
Bureau of Labor Statistics (BLS). The QCEW provides county-level
summaries of a variety of economic indicators, including employment,
number of firms, and average annual pay, broken down into industry and
sector. In order to match up with the National Broadband Map data,
annual average survey data released between 2011 and 2013 will be used.
Combining the National Broadband Map data with the QCEW data
results in a dataset that contains 3,142 counties with 6 observations per
county. As shown in Table 2, between the first and last time period,
roughly two-thirds of the counties experienced an increase in access to
fiber internet. Counties with an increase in access to fiber experienced
substantially more employment growth than counties that did not, and also
Spring 2015
16
had greater changes in the number of total firms and total average weekly
pay. This provides some initial evidence of a positive relationship between
access to fiber internet and employment growth; however, a simple
difference in means comparison is not sufficient to draw policy
conclusions from. There could be a variety of factors that contribute to
both job growth and improved internet infrastructure. Further, different
counties saw drastically different changes in internet access and
employment growth.
Table 3 breaks up the counties that experienced an increase in
access into quartiles (based on percent of households with access to fiber).
The relationship between the magnitude of the increase in access and the
change in economic indicators is more complex than the binary
comparison, although there are some indicators where there is clearly a
positive correlation, such as number of total firms. This provides evidence
for using a continuous rather than discrete or binary variable for access to
fiber internet.
Table 2. Comparing Economic Indicator Changes by Changes in Access to Fiber
between June 201 1 and December 2013
Source: National Broadband Map
Methodology
This study uses a two-way fixed effects regression1 to evaluate the
relationship between access to fiber internet connections and economic
growth. The model has fixed effects for county and for time. A fixed
effects regression is superior to a simple cross-sectional model or a pooled
ordinary least squares model in these circumstances because it allows the
model to control for unmeasured characteristics of counties that may be
Washington Academy of Sciences
17
correlated with access to fiber technology and influence measures of
economic growth in addition to factors that were common across all
counties for any given time period. If the hypothesis holds, counties that
experience increases in access to fiber internet will have greater increases
in employment than counties that have no change in high speed internet.
While the fixed effects model will not definitively prove causality, it does
provide a stronger case for causality than a cross-sectional model
(Whitacre et al., 2013).
Table 3. Comparing Economic Indicator Changes between 2011 and 2013 for Counties
that Increased Fiber Access
Source: National Broadband Map
The independent variable of interest will be percent of households
within a region that have access to fiber internet technology. Due to data
limitations, the percent of households having access serves as a proxy for
both individuals and businesses having access in that region. Because
GDP is not available at the county level, the primary dependent variable
will be employment, which is available. Additionally, the number of firms
and average annual wages will be used in order to provide a more
comprehensive overview of the economic impact. For both employment
and firms, natural logs will be used so that the results are meaningful
across counties of drastically different sizes. By running each model for
both the private sector and total economy, fiber internet’s impact on the
Spring 2015
18
private sector and the public and non-profit sector can be contrasted.
Control variables for county demographics and access to cable internet are
included to isolate the relationship between fiber internet and employment.
Exhibit 1 shows the model and variables that are the main focus of
this paper.
Exhibit 1. Model Variables and Predicted Relationships
Yit ~ Po + PiXut + P2X2U + P3X3U + P4X4U + Ps^sit + ai + at
where:
Washington Academy of Sciences
19
The model will examine the relationship at a county level; Kolko
(2011) showed that a state-level model is too aggregated to show
statistically significant differences in access to broadband. While there is
substantial variation in change in access to fiber internet at the state-level,
the small sample size and fact that most states are clumped at the lower
end of the spectrum would likely lead to a similar finding in this dataset.
Table 4 presents state level fiber optic data. Figures 1 and 2 illustrate that
there is much variation, at a county-level, in the level of access to fiber
technology, providing a robust dataset on which to conduct analysis.
Further, there is very little geographical concentration to where fiber is
being deployed, which will allow the results of this model to be applied
across all of the United States.
The National Broadband Map began data collection in 2010;
however, there were concerns over the quality of the first year’s data
collection methodology; the data were cleaned up for subsequent years
(Whitacre et al., 2013). Therefore, this study will examine data from each
of the releases in 2011, 2012 and 2013. While a larger dataset would be
ideal in order to understand the lasting effect of increasing access to fiber
internet, available data are sufficient to provide early evidence on the
relationship between access to fiber internet and economic indicators.
Policymakers will not delay actions for the next few years in order to
collect more data; neither should researchers.
Results
The results for the primary dependent variable, total employment,
are displayed in Table 5. Column (1) shows a simple one-way fixed
effects model with no control variables; the coefficient on access to fiber
internet is highly statistically significant, with a t-statistic of over ten.
When the natural log of population is controlled for in column (2), the
coefficient and its significance do not change substantially; the R squared
value rises from .017 to .949, though. This is as expected, as the
overwhelming determinant of how many employed people are in a county
will be population. In column (3), controls for changes in demographic
characteristics are added in. While the inherent wealth and education of
each county are absorbed by the unit fixed effects, adding these variables
accounts for any changes in income and education level that may have
occurred over the time period studied. We see that both of these controls
are statistically significant, as we would expect since wealth and education
are traditionally positively correlated with employment.
Spring 2015
20
Table 4. Percent of Households with Access to Fiber by State, Ranked by Access to
Fiber in December 2013, in both June 201 1 and December 2013
Washington Academy of Sciences
21
Figure 1.
Access to 1 gig/sec Download Speed by County (2011)
Source: National Broadband Map
Figure 2.
Access to 1 gig/sec Download Speed by County (201S)
Source: National Broadband Map
Spring 2015
22
Column (4) adds in a control for access to cable internet. This
ensures that any association between access to fiber internet and
employment growth is not actually due to the relationship between
employment and increased access to internet in general. The coefficient on
cable is surprisingly not statistically significant. Based on the body of
literature, a positive and statistically significant coefficient on access to
cable internet was anticipated. A possible explanation for this could be
that during the time period in question, roughly $5 billion in stimulus
funding was spent on expanding broadband access, much of which was
spent on expanding access to cable internet in rural areas of the country.
These areas that did not already have access to broadband likely were
some of the hardest hit and last to recover from the recession, explaining
why they lag behind in employment growth while experiencing an
increase in access to cable internet.
Columns (5) and (6) add in time-fixed effects. This is particularly
important as the country was recovering from the Great Recession during
this time, so employment growth could be the result of a generally positive
economic trend. The time fixed effects may account for part of the
coefficient for access to fiber internet, yet this coefficient is still
statistically significant at a 99 percent confidence level. Finally, column
(6) adds in robust standard errors to control for potential
heteroscedasticity. A control variable for state level stimulus spending
delivered through the National Telecommunications and Information
Association (NTIA) was also used, although not shown. Adding in the
control for NTIA stimulus spending had almost no impact to any of the
other coefficients, perhaps because the only available data are not at the
county level or accurate enough in terms of timing of implementation.
Similar models were run for the other dependent variables of
interest: total establishment count, total average weekly wages, private
sector employment, private establishment count, and private average
weekly wages; the results for model (6) are shown in Table 6. For the
wage models, median household income is replaced by the log of total
employment. Statistically significant coefficients are found for total and
private employment and total and private establishment count. The
coefficients on average weekly wages were significant until time-fixed
effects were added in, which soaked up most of the coefficient and
significance. Since the wages are in nominal values, the relationship
depicted prior to adding time-fixed effects was likely due to inflation.2
Washington Academy of Sciences
23
Table 5. Regression Results for Total Employment
t-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Overall, the results show evidence of a strong positive correlation
between the percent of households that have access to optical fiber internet
in a county and the number of employed individuals and number of firms.
Specifically, a 10 percent increase in the percent of households with
access to fiber internet is associated with a 0.13 percent increase in total
employment and a 0.1 percent increase in the number of firms. There is no
evidence of a relationship between access to fiber internet and average
weekly wages within a county.
Spring 2015
24
Table 6. Coefficient on Percent of Households with Access to Fiber Internet for each
Dependent Variable*
*Each model controls for 2- way fixed effects (county and date) and demographics.
Without a controlled or quasi-experiment, a causal relationship
between access to fiber internet and employment growth cannot be
claimed, but the results shown do support the theory that installing fiber
internet can help job growth. While controlling for time and unit fixed
effects and other controls helps to isolate the relationship between access
to fiber and employment growth, there is still the possibility that there are
unmeasured factors that influence both access to fiber and job growth.
While state level NTIA stimulus spending is controlled for, county level
spending cannot be controlled for due to data limitations. This creates a
slight problem; while the source of funding for the increase in access to
fiber is not the topic of this paper, stimulus funds had a specific goal of
creating jobs and contractors typically had to lay out a plan for hiring
additional employees as a part of their bid for stimulus funding. Therefore,
if some of the infrastructure that led to the increase in fiber access was
because of stimulus spending, it may have created more jobs than private
investment, which does not have to meet any job creation criteria. While a
better control for this would be ideal, it is unlikely that this is the primary
cause of the positive relationship. As mentioned previously, most of the
broadband stimulus spending went to expanding access to cable
technologies, not optical fiber.
Additionally, there is a possibility that job growth is driving access
to fiber internet, rather than the other way around. The positive
relationship could be due to internet service providers expanding into
growing areas. While it is likely that some of the positive relationship can
be attributed to this, it is unlikely to be the primary reason. Most of
America still is without access to fiber internet, so service providers would
be more likely to invest in areas where they already see demand rather
than trying to predict where employment growth will be. Additionally,
Washington Academy of Sciences
25
laying the infrastructure for fiber internet takes time and planning; since
this model looks at six month intervals, it is unlikely that service providers
saw employment growth in an area and were able to move in and offer
fiber service within six months.
A final critique of the model could be the relatively short-term
time frame used. Policymakers are not concerned with much longer time
frames than two years when investing heavily in internet technologies.
Unfortunately, the relative newness of the National Broadband Map data
set, and limitations of previous data collection efforts, limit the years that
can be examined. As data collection efforts continue, researchers should
continue to evaluate this relationship to test whether or not better internet
leads to sustained growth, or if growth is merely temporary.
Conclusion: Policy Relevance
This study provides evidence that increasing access to state of the
art internet like optical fiber and employment growth are related.
Policymakers considering investments in improving internet technologies
might consider these results when debating whether or not the cost of the
investment is appropriate. This information is useful to policymakers at all
levels of government, who have taken a variety of approaches to
improving access to ultra-high speed internet networks.
In January of 2015, the FCC changed its definition of broadband
internet from offering download speeds of 4 Mbit/sec or greater to
offering much faster download speeds of 25 Mbit/sec or greater. This was
a highly contentious shift in policy that will impact how data are collected
and what networks qualify for future public investments. Additionally, it
may change how the FCC views the state of competition within the
telecommunications industry, which could lead to other legislative,
executive or even judicial actions (Brodkin, 2014). While this study does
not address whether or not 25 Mbit/sec internet fosters more economic
growth than 4 Mbit/sec internet, it does provide preliminary evidence that
there could be a public interest in promoting faster internet speeds. This
contradicts what many of the detractors of the FCC’s change in definition
have argued; that the internet is fast enough and people do not benefit any
more from speeds over 25 Mbit/sec than they would at lower levels.
Another contentious policy area has been the recent development
of local (partially or fully) tax-payer funded high speed fiber networks
which offer internet speeds of up to 1 Gbit/sec (O’Toole, 2014). In
response to these networks, some states have considered blocking these
Spring 2015
26
efforts in order to prevent municipalities from crowding out private
expansion into high-speed internet markets. This paper does not provide a
cost benefit analysis of publicly-owned fiber optic networks, but does
provide evidence that policymakers should consider when deciding
whether or not these municipal fiber networks are wise uses of taxpayer
funds. On the other hand, though, the non-significant coefficient on access
to cable internet may provide evidence that pushing internet technologies
into underserved regions may not unilaterally lead to economic growth. A
more thorough examination of the specific investments made during the
stimulus act would provide better insight into this, though, as that was not
the primary focus of this paper.
As part of the American Recovery and Reinvestment Act, the FCC
developed the National Broadband Plan which outlines goals for internet
infrastructure in America. In the plan, the FCC set ambitious long-term
goals including providing affordable access to internet with speeds of 100
Mbit/sec or greater to at least 100 million homes and eventually ensuring
that every American has access to affordable fiber internet. The results of
this paper show that these are not unfounded goals, and there may be a
public, economic interest in achieving the goals outlined in the National
Broadband Plan.
Endnotes
A two-way fixed effects model controls for unmeasured variables that remain
constant throughout the time period for each county, as well as variables that are common
across all units for a single time period. These variables could potentially cause bias if
left uncontrolled for.
2 Diagnostic tests indicated that fixed effects are preferable to random effects and
suggested that robust standard errors are needed due to potential heteroscedasticity.
References
Brodkin, J. (2014, September 8). AT&T and Verizon say 10Mbps is too
fast for ‘Broadband’.
Crandall, R. W., Lehr, W., and Litan, R. E. (2007). The effects of
broadband deployment on output and employment: a cross-
sectional analysis of U.S. data. Issues in Economic Policy , 6
Washington, D.C.: Brookings Institution.
Washington Academy of Sciences
27
Czemich, N., Falck, O., Kretschmer, T., and Woessmann, L. (201 1).
Broadband Infrastructure and Economic Growth. The Economic
Journal, 727(552), 505-532.
Ezell, S. J., Atkinson, R. D., Castro, D., and Ou, G. (2009). The need for
speed: the importance of next-generation broadband networks.
Available at SSRN 1354032.
Feser, E., Horrigan, J., and Lehr, W. (2013, March). Symposium Report:
Findings from the Research Roundtable on the Economic and
Community Impact of Broadband. NC Broadband.
Fixed and wireless broadband subscriptions per 100 inhabitants. (Dec.
2013). OECD Broadband Portal. Retrieved on October 13, 2014
from http ://www . oecd. or g/ sti/broadband/ oecdbroadb andportal . htm .
Kelly, J. (2010, April 16). Next steps for our experimental fiber network.
Official Google Blog. Retrieved October 13, 2014.
Kolko, J. (2010). A new measure of U.S. residential broadband
availability. Telecommunications Policy, 34(3), 132-143.
Kolko, J. (2012). Broadband and local growth. Journal of Urban
Economics, 77(1), 100-113.
Koutroumpis, P. (2009). The economic impact of broadband on growth: A
simultaneous approach. Telecommunications Policy, 33(9), 471 -
485.
Lehr, W. H., Osorio, C. A., Gillett, S. E., and Sirbu, M. A. (2005).
Measuring broadband’s economic impact. In Broadband
Properties, December 2005, 12-24.
National Broadband Plan. (2010, March 17).
O’Toole, J. (2014, May 20). Chattanooga’s super-fast, publicly-owned
Internet.
http://monev.cnn.com/2014/05/20/technology/innovation/chattano
oga-internet/index.html
Spring 2015
28
Qiang, C. and Rossotto, C. (2009). Economic Impacts of Broadband. In
2009 Information and Communications for Development:
Extending Reach and Increasing Impact, (pp. 35-50). Washington,
D.C.: World Bank.
Selyukh, A. (2014, August 5). U.S. FCC asks if broadband should mean
faster Internet speeds.
http://www.reuters.com/article/2014/08/05/usa-intemet-speed-fcc-
idUSL2N0QB15S20 140805
Shideler, D., Badasyan, N., and Taylor, L. (2007, August). The economic
impact of broadband deployment in Kentucky. Federal Resewe
Bank of St. Louis Regional Economic Development \ 3(2), 88-1 18.
Stenberg, P., Morehart, M., and Cromartie, J. (2009). Broadband internet
service helping create a rural digital economy. Amber Waves, 7(3),
22-26.
Van Gaasbeck, K. A. (2008). A rising tide: Measuring the economic
effects of broadband use across California. The Social Science
Journal, 45(4), 691-699.
Whitacre, B., Gallardo, R., and Strover, S. (2013). Rural broadband
availability and adoption: Evidence, policy challenges, and
options. Webinar conducted for the National Agricultural and
Rural Development Policy Center, March 18, 2013.
Bio
Paul Lapointe is a recent graduate of the Masters of Public Policy
program at Georgetown University’s McCourt School of Public Policy,
where he focused his studies on domestic economic policy. Prior to
Georgetown, he worked in distribution and logistics analytics in the
private sector. Paul received his bachelor’s from the Robert H. Smith
School of Business at the University of Maryland.
The author may be contacted at plapoint50@gmail.com. The
diagnostics for this study are available to readers upon request.
Washington Academy of Sciences
29
Benjamin Banneker and Celestial Navigation:
Just How Did They Know Where They Were,
Then?1
Sethanne Howard
USNO, Retired
Abstract
Benjamin Banneker was an American scientist of the late eighteenth
century. He was a self-educated free black and became an expert in
astronomy, mathematics, and surveying. Major Andrew Ellicott asked
him to join the team surveying the original boundaries that became
Washington D.C. This paper presents Banneker’s story — which is
inspiring for all those who struggle against strong odds — and also
discusses the techniques used in those days to determine latitude and
longitude for surveying.
Introduction
Benjamin Banneker was bom in 1731 in Baltimore County, Maryland.
He died in 1 806 in Baltimore County, Maryland. He lived his entire life on
the family’s 100-acre tobacco farm near Oella, Maryland, a small hamlet
which is near Catonsville, Maryland. His mother, Mary, was a free black,
his father, Robert, was a freed slave from Guinea. Figure 1 shows a
reconstmction of his log cabin.
Figure 1. A reconstruction of Banneker’s cabin.
1 Presented at the Washington Academy of Sciences 2015 Annual Meeting and Awards
Banquet, May 14, 2015.
Spring 2015
30
One might think that free blacks were extremely rare. That is
almost true. The state of Maryland had the largest number of free blacks of
any of the states according to the 1830 census. There were over 52,000
free blacks in Maryland at that time.
Figure 2 shows a woodcut of Banneker. It probably is somewhat
idealized. It appeared on the cover of one of his publications, and in those
days, publishers felt free to embellish their publications.
Figure 2. Woodcut of Benjamin Banneker. Over the years the name Banneker has had
various spellings, and currently it is Banneker.
Washington Academy of Sciences
31
Banneker may have attended a few years of a nearby Quaker
school; however, the vast majority of his learning was self-taught. He
borrowed books from neighbors and studied them thoroughly. He loved
mathematics and astronomy and became more than proficient in them; he
was a skilled researcher, the equal of any of his contemporaries. Later in
life he expressed that,
“ The colour of the skin is in no way connected with
strength of the mind or intellectual powers .”
At age 22, Banneker built a wooden clock that struck the hours
throughout his life. Clocks were not common items in the late 1700s.
Local people came to marvel at his remarkable clock. He was not the first
person to build a clock in the colonies, but he was one of the rare few who
succeeded in doing so.
Banneker believed in seeking peaceful resolutions to conflicts.
Along with Dr. Benjamin Rush in 1792, he wrote a proposal to the Federal
Government asking the government to establish a Peace Office with equal
status to the Department of War. Almost 200 years later, the government
set up the United States Institute of Peace. The U. S. Institute of Peace
(USIP) works to prevent, mitigate, and resolve violent conflict around the
world. USIP does this by engaging directly in conflict zones and by
providing analysis, education, and resources to those working for peace.
Created by Congress in 1984 as an independent, nonpartisan, federally-
funded organization, USIP has more than 300 staff working at the
Institute’s Washington, D.C. headquarters and on the ground in the
world’s most dangerous regions.
A Bit of Mathematics
Benjamin Banneker loved math. He taught himself algebra,
geometry, trigonometry, and spherical trigonometry. He drew great delight
from creating math puzzles and solving them. One of his puzzles is:
Divide 60 into four such parts that the first being increased
by 4, the second decreased by 4, the third multiplied by 4,
the fourth divided by 4 such that the sum, the difference,
the product, and the quotient shall be one and the same
number.2
2 The answer to the puzzle appears at the end of this paper.
Spring 2015
32
A Bit of Astronomy
Banneker liked to lie outside his log cabin all night watching the
sky. So far as we know he did not own a telescope, although he knew how
to use one (his neighbors, the Ellicotts, had telescopes). Those math skills
he had learned he applied to astronomy. In 1788, he accurately predicted
the solar eclipse of 1789. Predicting eclipses had long been the province of
professional astronomers. The techniques for doing so were not commonly
taught. He timed the eclipses of the Galilean satellites by Jupiter. Figure 3
shows those satellites with Jupiter.
Callisto
Ganymede
Europa
lo
Jupiter
Figure 3. The Galilean satellites of Jupiter.
Surveying
President George Washington asked Major Andrew Ellicott to
survey the land we now call Washington, D.C. It became the District of
Columbia in 1801. A Commission was set up to oversee the project. In
1791, Ellicott asked Banneker to be part of the survey team. Banneker was
59 when he joined Ellicott’s team. He was hired for his astronomical and
mathematical knowledge. So he spent several months slogging through the
swampy land putting down milestone markers.
Washington Academy of Sciences
33
The original plan for Washington, D.C., called for 10 miles on a
side square, equaling 100 square miles of land. For every mile of the
perimeter, the survey team laid down boundary markers. Figure 4 shows
the original boundary marker at 6980 Maple Street, N.W. Many of the
boundary markers have disappeared over the years, but a few remain.
Some of those have Banneker’s name on them.
Figure 4. Boundary Marker on Maple Street, N.W.
In 1846, Washington, D.C., gave the Virginia portion of the
District of Columbia back to Virginia, leaving the District we have today.
George Washington asked l’Enfant to design the new city.
F Enfant drew up a plan for the city (see Figure 5), but ran afoul of the
Commission overseeing the project. He left the project. Major Ellicott then
drew up a city plan (see Figure 6) that was used to construct the original
city.
Spring 2015
34
Figure 5. l’Enfant’s plan for the city.
I.
\M v
i
111!
jlltljill
r* _• w «
5 ji
4 jj
ii
|1|
Hi i
|}]Li
=! h -m
Its! i
Washington Academy of Sciences
jV.I/,S OA /‘o/.K.V
35
Note that Major Ellicott set 0° longitude at the Capitol Building.
The 0° longitude meridian was set by many nations at many places over
the centuries. In most cases, however, the meridian at Greenwich,
England, was used as 0° longitude because England ruled the seas in the
eighteenth century.
It was not until 1884 by international treaty that Greenwich was
chosen as the permanent 0° longitude — the Prime Meridian.
VI. AN
.' rL ifiiq ■* *Wdlh»iMl0n -
in i hr Trrrilory <4 CpIubiIma.
— .w-« V M ■ -i __
I '//(G/.s l. / --T/ . »/. i/ti'L. f \ fl
llnurt itmrruc
MB V. tVr. rft.h
. V/;. 'tiRXMHXT.
.</t. ■ t* . \\
Si Din . ’
■^ErrBF.FFfF^ |:.-3
[t [ti— ir- p rpp-^rprp ticsrr t“
,>rrF K“ r^EEEF ZeEMF'EFJL'.F-
— rrr r.-srs [■ — n ■ i — p^-
FTIRFEWEV -■ t /UFERFFFFFI rZF
,, FFFFr^-F &■*< r^'~F1FFF0PfiFFP.r'!^r
CiKORGR Town. .^njp^pri^BkEB FIF .^nUFRPEFPrj.r- F~
■7nF«^3iHFH~iE>np p’fB^crFF.^-r—
.. ( \ -Efc igBBEJBEEte.'t—JsSfTgBBCgEEEEfe ‘ ’ ^
✓ ■■ i rr v4Ck n-rr-r^ raui rrr k i
/ ■ trctEferr-RRFr ^ «■ Ere^liBewHrwi f*i
t>. v\ : i^EEEFf.: i afig^v r- ^CRFEFaSRSfera,
jObfmuuiono '>■
ocpliiniilorv of llir plan.
A\V-
JjrraStli .if tin $»irrris.
Figure 6. Major Ellicott’s city plan.
The Nitty-Gritty of 1791 Surveying
To survey is to measure the latitude and longitude of the perimeter
of the land under consideration. Today surveyors use GPS, but in 1791
GPS did not exist. Surveyors turned to events in the sky to determine
latitude and longitude. This means astronomy.
Spring 2015
36
It takes two coordinates to find a location on a sphere [we do all
agree that we live on the surface of a sphere, the Earth]: a left/right one
and an up/down one. Figure 7 illustrates the concept.
Figure 7. The concept of latitude and longitude.
The up/down is latitude which runs from 0° at the equator to ±90°
at the poles. The left/right is longitude which runs from 0° to 180° East
and 0° to 180° West. We call the latitudes circles of latitude ; we call the
longitudes meridians of longitude. See Figure 8 for an illustration.
Latitude
North
90 (+) 30
South
(-)
Longitude
0
Prime
meridian
Figure 8. Illustration of the circles of latitude and meridians of longitude.
Washington Academy of Sciences
37
Let us start with latitude. People knew how to find their latitude by
the 2nd century BCE. Start with a sextant and sight along your horizon. In
the northern hemisphere then find Polaris, the North Star — no telescope
needed. Measure the altitude of Polaris (how far you tilt upward with your
sextant) with the sextant. Your latitude is 90° minus the altitude of Polaris.
It is a simple procedure. Figure 9 illustrates the concept. Latitude uses
angles ranging from 0° at the equator to ±90° at the poles.
zenith
S
N
altitude
nadir
Figure 9. How to measure the altitude of Polaris.
Spring 201 5
38
Longitude, on the other hand, is not so simple. Someone must
define a starting point for 0° longitude. The British colonies (and the new
United States) used Greenwich as the Prime Meridian. The time at
Greenwich is known as Greenwich Mean Time (GMT) or Universal Time
(UT).
Longitude uses both time and angle. That begs the question of how
longitude relates time to angle. The Earth is not sitting still; it is spinning
on its axis. This does not matter for measuring latitude. It does matter for
longitude. The Earth rotates once around each day, or 360°/day. A day has
24hrs. Work your way down from this to the Earth rotates 15° in lhr. So if
you are at the equator, you are spinning with the Earth at 1675 km/hr. If
you can measure your local time and the time at Greenwich, you can get
your longitude. For example, if it is 1800hrs at Greenwich when it is 9h20m
on the same day, then your local time is 8h and 40m behind Greenwich
Time. Convert that value to angles, and you have your longitude:
Long = -
(8x15°)
+
( 40
— xl5°
Uo
y
= -[120° + 10°] = -130° = 130° West
However in the 1700s, one could not call Greenwich from
Washington, D.C., and ask the time. So the survey team needed to use
events in the sky. They did have an ephemeris. An ephemeris is a time-
ordered list of future positions of the Sun, Moon, planets, satellites, and
stars (at 0hrs GMT). Astronomers were paid to develop ephemerides for
various places. This took a great deal of mathematical skill. Most travelers
took an ephemeris with them on their travels.
An Almanac contains an ephemeris along with other important
information such as holidays, eclipses, sunrise, and sunset. The Federal
Government still publishes the Astronomical Almanac and the Nautical
Almanac each year. An example of an ephemeris for Mars for the month
of June 2015 looks like this:
Target body
Start time
Stop time
Step-size
Date (UT)
name: Mars
: 2015-May-31 00:00:00.0000 UT
: 2015-Jun-30 00:00:00.0000 UT
: 1440 minutes
HR:MN R.A. DEC
Washington Academy of Sciences
39
^ •X' *1* *1* *1* <1* «L* si# si* vl/ si* si* si* si* si* si* si* si* sis si* si,* si* si* si* si* si* si* i* si* si* si* si* si* ^L* si* si*
'I' *x* *T* 'T* 'T' /T' -T* *T* 'T* 'T* 'T' ^ #ys #ys #ys *Js *Js *Js *Js ^ *ys *Js *Js *ys *Js *Js *Js *Js *Js *|S *Js *Js *[S *ys #*jv *|S *js *[S 'p 'P
2015-May-31 00:00
2015-Jun-01 00:00
2015-Jun-02 00:00
2015-Jun-03 00:00
2015-Jun-04 00:00
2015-Jun-05 00:00
2015-Jun-06 00:00
2015-Jun-07 00:00
2015-Jun-08 00:00
2015-Jun-09 00:00
2015-Jun-10 00:00
2015-Jun-l 1 00:00
2015-Jun-12 00:00
2015-Jun-13 00:00
2015-Jun- 14 00:00
2015-Jun-15 00:00
2015-Jun-16 00:00
2015-Jun-17 00:00
2015-Jun- 18 00:00
2015-Jun-19 00:00
2015-Jun-20 00:00
2015-Jun-21 00:00
2015-Jun-22 00:00
2015-Jun-23 00:00
2015-Jun-24 00:00
2015-Jun-25 00:00
2015-Jun-26 00:00
2015-Jun-27 00:00
2015-Jun-28 00:00
2015-Jun-29 00:00
2015-Jun-30 00:00
The coordinates are right ascension and declination — two common
astronomical coordinates.
There are two important clocks one needs for surveying in the
1700s: a clock keeping GMT and a clock keeping local time. If the GMT
clock stops, it is very difficult to retrieve the correct GMT (one cannot call
Spring 2015
40
home). If the local clock stops, it is not as difficult to retrieve the local
time but it does take some effort.
Banneker had the responsibility for keeping the clocks wound.
This was a vital position to have.
Now they needed a celestial event to time. There were a few
schemes used in the 1700s — most were not very precise. One celestial
event that showed promise was the planet Jupiter eclipsing the four
Galilean satellites (the four brightest moons of Jupiter, discovered by
Galileo), see Figure 3. They had an ephemeris for the eclipse times for
these satellites. Today you can use a smart phone app to get the ephemeris
for the Galilean satellites.
Both the Mason-Dixon Line and the boundary for Washington,
D.C., were set using the eclipses of the Galilean satellites.
Banneker worked on the survey team for just a few months. Ill
health drove him back to his farm. He continued to compute ephemerides
and, beginning in 1792, published a series of six Almanacs. They sold in
six cities in four states for the years 1792 through 1797: Baltimore;
Philadelphia, Pennsylvania; Wilmington, Delaware; Alexandria, Virginia;
Petersburg, Virginia; and Richmond, Virginia. They were best sellers at
the time. Today, very few exist. The Maryland Historical Society has one.
The cover for the 1792 edition is shown in Figure 10.
People who did not have clocks depended on an Almanac to give
them sunrise and sunset times so they could tell the time of day.
Summary
Benjamin Banneker was an American scientist of repute. As a
testament to his reputation, the Federal Gazette wrote the following
obituary: “Mr. Banneker is a prominent instance to prove that a
descendant of Africa is susceptible of as great mental improvement and
deep knowledge into the mysteries of nature as that of any other nation.”
There are some who say that his intellect matched that of Ben Franklin.
There are many schools named after Banneker, and the Benjamin
Banneker Museum and Park is maintained by Baltimore Recreation and
Parks. Its address is 300 Oella Avenue, Catonsville, MD 21228.
Washington Academy of Sciences
'MMHI
41
r
Benjamin Banneker’s
PENNSYLVANIA, DELAWARE,
MARYLAND and VIRGINIA
EPHEMERIS,
For the YEAR of our LORD,
1792;
Being BISSEXTILE, or LEAP-YEAR, and the Six-
teenth Year of AMERICAN INDEPENDENCE,
which commenced July 4, 1776.
Containing, the Motions of the Sun and Moon, thetrut
Places and AfpcCts of the Planets, the Riling and Setting of
the Sun, and the Riling, Setting and Southing, Place and Age
of the Moon, See. — The Lunations, Conjunctions, Eclipfes,
Judgment of the Weather, Feftivals, and other remarkable
Days ; Days for holding the Supreme and Circuit Courts of tin
United States , as alfo the ufual Courts in Pennfylvania , Dela-
ware, Maryland, and Virginia. — Also, feveral ufcful Tables,
and valuable Receipts. — Various Selections from the Com-
monplace-Book ot the Kentucky P Lilcf ] her , an American Sage j
with interefting and entertaining Elfays, in Profe and Verfc —
the whole comprifing a greater, more pleafmg, and ufeful Va
riety, than any Work of the Kind and Price in North- America.
BALTIMORE: Printed and Sold, Wholefale and Retail, b>
William Goddar d and James Angell, at their Print-
ing-Office, in Market- Street. — Sold, a!fo, by Mr. Joseph
Crukshank, Printer, in Market-Street , and Mr. Daniei'
Humphreys, Printer, in Soutb-Frcnt-Street, Philadelphia
and by MefTrs. Hanson and Bond, Printers, in Alexandria I
Figure 10. 1792 cover for the Banneker Almanac.
Spring 2015
42
Bio
Sethanne Howard is an astronomer and retired Chief of the
Nautical Almanac Office at the U.S. Naval Observatory. She maintains
her research field of interacting galaxies. As the first woman to receive a
bachelor’s degree in physics from the University of California, Davis, she
went on to get a master’s degree in nuclear physics from Rensselaer
Polytechnic Institute, and a PhD in astrophysics from Georgia State
University. She worked at several astronomical observatories, at NASA
managing operational astrophysical satellites, at NSF as Program Officer
for Extragalactic Astronomy and Cosmology, and finally at the U.S. Naval
Observatory.
The answer to the math puzzle presented
earlier in this paper is:
W = 5.6 is the first part
X = 13.6 is the second part
Y = 2.4 is the third part
Z = 38.4 is the fourth part
W + X + Y + Z = 60
W + 4 = X- 4 = Y*4 = Z/4 = 9.6
One needs to solve the set of simultaneous
equations to get the solution.
Washington Academy of Sciences
43
Washington Academy of Sciences
Awards Program 2015
Background
The purpose of the Washington Academy of Sciences, which was
founded more than a century ago in 1898, is to encourage the
advancement of science and “to conduct, endow, or assist investigation in
any department of science.” To recognize scientific work of distinction,
the Academy gives awards annually to scientists who work in the greater
Washington, D.C., area. The awards are presented by colleagues at the
academy’s annual business meeting and awards ceremony.1 The public is
invited to help celebrate and recognize the extraordinary achievements of
the honored scientists and engineers, so the Academy hosts a formal
Business and Awards Banquet in the Washington area. At this ceremony,
the nominating colleague gives a short 3 -minute introduction describing
the awardee, and the awardee must be present to accept the award, but the
tradition of requiring formal acceptance speeches ended back in 1955.
Photo: A1 Teich
Washington Academy of Sciences annual Awards Banquet at the conference center of the
National Rural Electric Cooperative Association (NRECA) in Arlington, Virginia, May
14, 2015.
1 Per the Academy’s by-laws, the annual business meeting takes place by the third
Thursday in May, and usually consists of brief reports by the outgoing and incoming
presidents and an audit report.
Spring 2015
44
Awards Program Early History
While the Academy passed its centennial year in 1998, the
Academy’s Awards Program has featured 75 years of achievement at this
point in time. It’s interesting to recount the history of the program which
began in 1940. The Academy’s Bylaws had been amended the previous
year to permit the Academy to award “medals and prizes . . . [for] scientific
work of high merit.”
At that time, 1939, the Academy’s Board of Managers established
awards for noteworthy accomplishments during the year by young
scientists — no more than 40 years old — in the biological, physical, and
engineering sciences. A proposal to raise the age limit to 45 for the
Biological, Engineering, and Physical Sciences categories was rejected in
1953. The requirement that “candidates shall not have passed their 41st
birthday” was dropped later on in 1982, and 1983 was the first year in
which an award for a Distinguished Career in Science was given.
Some Award Traditions
The year 1956 marked the first year that more than one award was
presented in a given category. In 1961, the Board of Managers officially
encouraged granting more than one award in any given category should
multiple qualified candidates exist.
Traditionally, the Academy’s awards have been given for work
done in the Washington, D.C., area. Since the Washington Academy of
Sciences was incorporated in 1898, the year 1998 marked the Academy’s
centennial year and the D.C.-area tradition was waived during that year —
as some awards were given to individuals affiliated with organizations
outside the D.C. area." In 1998, the Academy gave fourteen Centennial
Awards for Lifetime Achievement in Science, including awards in Science
Policy, Technology Policy, and History of Science. The next year, 1999,
was the first non-centennial year in which the award for Science Policy
was given. The History of Science award was not given again until 2012,
which was the same year Service to Science was first awarded. An award
for Lifetime Achievement in the Public Understanding of Science was first
made in 2014.
2 The Washington Academy of Sciences founders included Alexander Graham Bell and
Samuel Langley, Secretary of the Smithsonian Institution from 1887 to 1906.
Washington Academy of Sciences
45
Establishment of the Education and Teaching Awards
The first year an award was given for the Teaching of Science was
1952. For this award category, and for this category only, the age
limitation of 40 years was waived. This “special award” was given in 1952
and 1953. The award category for the Teaching of Science was officially
established in 1956.
In 1976, the Berenice Lamberton Award for the Teaching of
Science in High Schools was established. Lamberton was a professor at
Georgetown University with a long-time interest in education. The
Washington Academy of Sciences’ Junior Academy was initially set up by
Lamberton and others at Georgetown.
The Leo Schubert Award for College Teaching was established in
1979. The year 2000 was the first year in which awards were given for
Achievement in Education and Teaching of Science in Middle Schools.
In 2002, the Board of Managers, acting on the recommendation of
the Awards Committee, established the Marilyn Krupshaw Award for
Non-Traditional Education/Teaching. The award was named in honor of
the long-time leader of George Washington University’s Science and
Engineering Apprentice Program (SEAP) for high school students,
sponsored by the U.S. Department of Defense (DoD). The award was
presented for the first time in 2004. And in 2005, a special award was
given for Service to Science Education.
Nomination Process
The Academy welcomes nominations for its Awards Program each
year. The following is a complete list of the award categories as
established by the Academy’s Board of Managers:
• Distinguished Career in Science
• Biological Sciences
• Engineering Sciences
• Physical Sciences
• Health Sciences
• Behavioral and Social Sciences
• Mathematics and Computer Science
• Krupsaw Award for Non-Traditional Teaching
• Lamberton Award for Teaching of Science in High School
• Leo Schubert Award for Teaching of Science in College
Spring 2015
46
• Special Award ( e.g ., science policy, lifetime achievement in
education)
To carry out the Awards Program each year, the Academy’s Vice
President for Membership appoints an Awards Committee which sets the
submission dates for the year. Please watch the Academy’s website,
www.washacadsci.org, for those deadlines, typically in early Spring. To
nominate an individual, print and complete the Nomination Form that is
available at the website, and mail it directly to the Awards Committee as
indicated on the form.
The Awards Committee typically uses the standard categories that
appear on the nomination form, but when necessary, the Special Award
category may be used to include other categories. For example, an award
for Mathematics was first given in 1960. This other award category was
expanded to Mathematics and Computer Sciences in 1979. The Academy
first made an award for Behavioral Sciences in 1976; this other award
category was changed to Behavioral and Social Sciences in 1987. 4 Awards
were first given for two other categories — Health Sciences and
Environmental Science — in 1997, and later for Public Health. The year
2000 was the first year awards were given for the categories of
Anthropology and Astronomy. Back in 1961, the Board rejected a
proposal that an Earth Sciences award category be instituted; it was not
until 2014 that an award for Lifetime Achievement in Natural Resources
Sciences was made.
2015 Annual Banquet
At the Washington Academy of Sciences Annual Business and
Awards Banquet on May 14, 2015, the Academy’s ceremony honored an
illustrious group of individuals for their work in physical, biological, and
engineering sciences and other areas.
Ronald Colie received the 2015 award for Distinguished Career in
Science in recognition of his “lifetime work and major contributions in
radionuclidic metrology. Within the world of radioactivity measurements,
it is almost impossible to hear the words ‘radon,’ ‘uncertainty’ or
3 Since 1 990, additional awards have also been given at the annual awards ceremony for
special recognition of Service, or Meritorious Service, to the Washington Academy of
Sciences; however, these awards go through a different process.
4 Ainitai Etzioni, noted earlier in this issue of the Journal of the Washington Academy of
Sciences, was a recipient of this award in 1988.
Washington Academy of Sciences
47
‘metrologisf without thinking of the name Dr. Colie.” He is a specialist in
nuclear radiochemistry and the development of standards, and he and his
collaborators developed methods to analyze and standardize
brachytherapy sources, pellets of radioactive material designed to be
implanted in the body at sites requiring direct radiation exposure.
The Academy presented its 2015 award for Distinguished Career
in Engineering Sciences to Dr. Ram Duvvuru Sriram in recognition of
“contribution and technical leadership in developing computational tools
and techniques for engineering design and for enabling interoperability of
CAD/CAM/CAE systems.”
A 2015 award for Physical and Biological Sciences was presented
to Dr. Marcus Cicerone in recognition of “establishing and pioneering
the use of Broad Band Coherent Anti-Stokes Raman Spectroscopy
imaging and establishing exquisite optical techniques for examining the
dynamics of proteins and other biological molecules in the glassy sugar
matrices commonly used for their preservation.”
The Academy’s 2015 award for Biological Sciences was presented
to Dr. Paul M. Peterson in recognition of being a “tireless and prolific
taxonomist, collector, and publisher who has extensively revised the
classification of the large grass subfamily Chloridoideae and its genera,
and is leading the effort to prepare a DNA database for the grasses of
North America and noxious weeds for the Bar Code of Life.”
The Academy presented its 2015 award for Engineering Sciences
to Dr. Robert Gover in recognition of “work at the Naval Research
Laboratory on the development, implementation, and application of high-
fidelity physics-based digital models for the development of optimized
Electronic Warfare countermeasures against modem anti-shipping cruise
missiles.”
A 2015 award for Physical Sciences went to Mr. Gregory Strouse
in recognition of “international leadership in high-precision temperature
metrology, and innovative contributions to next-generation temperature
sensors.”
The Krupsaw Award for Non-Traditional Teaching was presented
in 2015 to Ms. MaryBeth Petrasek in recognition of her “teachings in the
techniques of medicolegal death investigation and forensic pathology to
young people.”
A 2015 award was also presented to Dr. Sally Rood in special
recognition of Service to the Academy for “momentous work as editor of
Spring 2015
48
the Journal of the Washington Academy of Sciences and coordination of
the agreement to begin the process of digitally preserving more than 100
years of the Journal’s published works.”
Photo: A1 Teich
Lisa Karan presenting the award for Distinguished Career in Science to Ronald Colie.
Photo: A1 Teich
Award for Distinguished Career in Engineering Sciences presented to Ram D. Sriram
(right) by Steven Fenves.
Washington Academy of Sciences
49
Photo: A1 Teich
Award for Physical and Biological Sciences, presented to Marcus Cicerone by Laurie
Locascio.
Photo: A1 Teich
Award for Biological Sciences, presented to Paul Peterson (left) by Chris Puttock.
Spring 2015
50
Photo: A1 Teich
Award for Engineering Sciences, presented to Robert Gover (left) by Douglas Fraedrich.
Award for Physical Sciences, presented to Gregory Strouse (right) by Gerald Fraser.
Washington Academy of Sciences
51
Photo: A1 Teich
Krupsaw Award for Non-Traditional Teaching, presented to MaryBeth Petrasek by Anne
Cupero (left).
Photo: A1 Teich
Special recognition for Service to the Washington Academy of Sciences was presented to
Sally Rood (right) by Master of Ceremonies Terrell Erickson.
Spring 2015
52
Washington Academy of Sciences
53
1 200 New York Ave.
Suite 113
Washington DC
20005
wvwv. wa shacadsci.org
Addendum* to
Washington Academy of Sciences
2014 Membership Directory
M=Member; F=Fellow; LF=Life Fellow; LM=Life Member;
EM=Emeritus Member; EF=Emeritus Fellow
Adkins, Michael K. (Mr.) 4143 Elizabeth Lane, Annandale VA 22003
(M)
Arif, Muhammad (Dr.) National Institute of Standards and Technology
(NIST), 100 Bureau Drive, MS 8460, Gaithersburg MD 20899-8460 (M)
Berry, Jesse F. (Mr.) 2601 Oakenshield Drive, Rockville MD 20854
(M)
Boisvert, Ronald F. (Dr.) National Institute of Standards and Technology
(NIST), 100 Bureau Drive, MS 8910, Gaithersburg, MD 20899-8910 (F)
Brown, Elise A. B. (Dr.) 681 1 Nesbitt Place, Mclean VA 22101-2133
(LF)
* These twenty names were inadvertently omitted from the Academy’s 2014
Membership Directory in the Winter 2014 issue of the Journal of the Washington
Academy of Sciences, so we are printing them here instead of waiting to include them in
the 2015 membership listing. As indicated in the inside cover of each quarterly Journal,
the last issue of the year contains a directory of the current membership of the Academy.
Spring 2015
54
Buford, Marilyn (Dr.) 3073 White Birch Court, Fairfax VA 22031 (F)
Caws, Peter J (Dr.) 2475 Virginia Avenue, NW, Apt. 230, Washington
DC 20037 (M)
Cupero, Jerri Anne (Dr.) 2860 Graham Road, Falls Church VA 22042
(F)
Danner, David L. (Dr.) 1364 Beverly Road, Suite 101, McLean VA
22101 (M)
Elster, Eric Andrew (Dr.) 3223 Geiger Avenue, Kensington MD 20895
(F)
Hollinshead, Ariel (Mrs.) 23465 Harbor View Road, #622, Punta Gorda
FL 33980-2162 (F)
Jayarao, Arundhati (Dr.) 881 1 Trafalgar Court, Springfield VA 22151
(M)
Kaufhold, John (Dr.) 4601 N. Fairfax Drive, Suite 1200, Arlington VA
22203 (M)
Martin, Charles R. (Dr.) P.O. Box 98521, M/S NLV085, Las Vegas NV
89193 (F)
Mittleman, Don (Dr.) 4650 54th Avenue S., Apt. 57B, St. Petersburg FL
33711-4638 (F)
O’Hare, John J. (Dr.) 108 Rutland Boulevard, West Palm Beach FL
33405-5057 (EF)
Sozer, Amanda (Dr.) 4707B Eisenhower Avenue, Alexandria VA
22304 (M)
Snieckus, Mary (Ms.) 1700 Dublin Drive, Silver Spring MD 20902 (M)
Washington Academy of Sciences
55
Williams, Jack (Dr.) 6022 Hardwick Place, Falls Church VA 22041 (F)
Williams, Tenisha (Ms.) 1209 7th Street, NW, Washington DC 20001
(M)
Wu, Keli (Mr.) 360 Swift Avenue, Suite 48, South San Francisco CA
94080 (M)
Spring 2015
56
Washington Academy of Sciences
57
In Memoriam
Burton G. Hurdle
(1918-2015)
Burton Garrison Hurdle, a Fellow of the Washington Academy
of Sciences, passed away peacefully on March 4, 2015. He was a research
physicist with the Naval Research Laboratory’s Acoustics Division
beginning in 1943 for 50 years.
Hurdle was bom in 1918 in Roanoke, Virginia, the son of Grover
Cleveland Hurdle and Bronna Rene (Garrison) Hurdle. He was raised
during the Great Depression and graduated from Jefferson Senior High
School in Roanoke in June 1936. After graduation from high school, he
went to work for the Norfolk and Western Railway that had its
headquarters in Roanoke. In that period, he started taking night classes and
then switched to becoming a full time undergraduate student at Roanoke
College.
In 1941, he received his B.S.
degree in physics, with a minor in
mathematics, and then enrolled at
Virginia Polytechnic Institute for
graduate studies. He intended to
major in mechanical engineering at
Virginia Tech, but after only two
weeks decided to major once again in
physics. While he was studying for
his Master’s Degree in physics he
taught some classes in the
university’s Mathematics Department
to supplement his income. He also
had an industrial fellowship with the
Standard Register Company of
Dayton, Ohio. Although he was
within about a year and a half from
receiving a Doctorate in Physics, he
left the university to join the U.S. Navy for the war effort, and so was
awarded a M.S. degree in General Physics at that time. While at Virginia
Tech, he had interviewed with recruiters from the Naval Research
Laboratory (NRL). After considering several other potential job
Spring 2015
58
opportunities, he accepted a position in the NRL Sound Division as a
Research Physicist and started work there in 1943. NRL was doing much
applied research then in support of the War effort. His doctoral thesis topic
was on the subject of acoustic interference fields in the ocean. Hurdle
worked under all five Superintendents of the Acoustics Division. His first
supervisor at NRL was Dr. Raymond Steinberger, and his early senior
NRL colleagues included Harvey Hayes, Raymond Steinberger, and
Prescott Arnold who were all Harvard-educated scientists.
Hurdle briefly left NRL during the period 1947-1949 to work at
Engineering Research Associates’ Physics and Chemistry Division in
Arlington, Virginia. During this period, he worked on several research
projects including investigations of the sound speed and absorption in
liquids using an interferometer; development of methods for calibration of
accelerometers using free-free bars; and development of methods for
calibrating acoustic pressure gauges and impulse gauges for use in
measuring the propagation of elastic energy in soil and rock.
Hurdle completed his Ph.D. at a later time, during work in the
United Kingdom.
In addition to being a Fellow of the Washington Academy of
Sciences, Dr. Hurdle was also a Fellow of the Acoustical Society of
America (ASA). He served the ASA in various capacities including the
Membership Committee, the Underwater Acoustics Technical Committee,
the Nominating Committee, and the Publications Policy Committee.
Dr. Hurdle was also a member of Sigma Xi. He served as
Associate Editor of the U.S. Navy’s Journal of Unden\>ater Acoustics
(1979-2004). He also served as General Chairman and Session Chairman
at meetings of the U.S. Navy Symposia on Underwater Acoustics. Dr.
Hurdle received numerous awards and commendations including the Alan
Berman Research Publication Award for “The Nordic Seas” in 1985 and
the Navy Superior Civilian Service Award in 1987. In 1998, he was the
recipient of the Distinguished Technical Achievement Award from the
Oceanic Engineering Society (OES) of the Institute of Electrical and
Electronic Engineers (IEEE). He was cited for his outstanding
contributions to understanding the oceanography and acoustics of the
Nordic Seas.
Washington Academy of Sciences
59
Washington Academy of Sciences
1200 New York Avenue, NW
Room 1 1 3
Washington, DC 20005
Membership Application
Please fill in the blanks and send your application to the Washington
Academy of Sciences at the address above. We will contact you as soon as
your application has been reviewed by the Membership Committee. Thank
you for your interest in the Washington Academy of Sciences.
(Dr. Mrs. Mr. Ms.)
Business Address
Home Address
Email
Phone
Cell
Please indicate:
Preferred mailing address
Type of membership
Business
Regular
Home
Student
Present Occupation or Professional Position
Please list memberships in scientific societies - and include office held:
Spring 2015
60
Journal of the Washington Academy of Sciences
Instructions to Authors
The Washington Academy of Sciences publishes its interdisciplinary peer-
reviewed Journal of the Washington Academy of Sciences four times a
year — Spring, Summer, Fall, and Winter.
1 . Deadlines for quarterly submissions are:
Spring - February 1 Fall - August 1
Summer - May 1 Winter - November 1
2. Draft manuscripts using a word processing program (such as
MSWord), not PDF.
3. Papers should be 6,000 words or fewer. With 7 or more graphics,
reduce the number of words by 500 for each graphic.
4. Include an abstract of 150-200 words.
5. Graphics must be in black & white or greytone. They must be
referenced in the text.
6. Use endnotes, not footnotes. The bibliography may be in a style
considered standard for the discipline or professional field
represented by the paper.
7. Submit papers as email attachments to the editor.
8. Include the author’s name, affiliation, and contact information —
including postal address. Membership in an Academy-affiliated
society may also be noted.
9. Manuscripts are peer reviewed and become the property of the
Washington Academy of Sciences.
10. There are no page charges.
Please see the Academy’s web site, www.washacadsci.org, for the library
subscription rate, listing of articles dating to 1899, and information on
accessing them.
Washington Academy of Sciences
61
Washington Academy of Sciences
Affiliated Institutions
National Institute for Standards & Technology (NIST)
Meadowlark Botanical Gardens
The John W. Kluge Center of the Library of Congress
Potomac Overlook Regional Park
Koshland Science Museum
American Registry of Pathology
Living Oceans Foundation
National Rural Electric Cooperative Association (NRECA)
Spring 20 1 5
62
Delegates to the Washington Academy of Sciences
Representing Affiliated Scientific Societies
Acoustical Society of America, Washington Chapter
American Association of Physics Teachers, Chesapeake Section
American Astronomical Society
American Ceramics Society
American Fisheries Society
American Institute of Aeronautics and Astronautics
American Meteorological Society, Washington, DC Chapter
American Nuclear Society, Washington DC Section
American Phytopathological Society, Potomac Division
American Society for Cybernetics
American Society for Metals, Washington Chapter
American Society of Civil Engineers, National Capital Section
American Society of Mechanical Engineers, Washington
Section
American Society of Microbiology, Washington Branch
American Society of Plant Biologists, Mid-Atlantic
Anthropological Society of Washington
ASM International
Association for Women in Science, DC Metropolitan Chapter
Association for Computing Machinery, DC Area Chapter
Association for Science, Technology, and Innovation
Association of Information Technology Professionals
Biological Society of Washington
Botanical Society of Washington
Capital Area Food Protection Association
Chemical Society of Washington
District of Columbia Institute of Chemists
District of Columbia Psychological Association
Eastern Sociological Society
Electrochemical Society, National Capital Section
Entomological Society of Washington
Geological Society of Washington
Historical Society of Washington DC
Human Factors and Ergonomics Society, Potomac Chapter
Institute of Electrical and Electronics Engineers, Northern
Virginia Section
Paul Arveson
Frank R. Haig, S. J.
Sethanne Howard
Vacant
Lee Benaka
David W. Brandt
Vacant
Charles Martin
Vacant
Stuart Umpleby
Vacant
Vacant
Daniel J. Vavrick
Vacant
Mark Holland
Vacant
Toni Marechaux
Jodi Wesemann
Alan Ford
F. Douglas
Witherspoon
Chuck Lowe
Stephen Gardiner
Chris Puttock
Keith Lempel
Elise Ann Brown
Vacant
Tony Jimenez
Ronald W.
Mandersheid
Vacant
Vacant
Jeffrey B. Plescia
Jurate Landwehr
Vacant
Gerald P. Krueger
Murty Polavarapu
(continued on next page)
Washington Academy of Sciences
Delegates to the Washington Academy of Sciences
Representing Affiliated Scientific Societies
(continued from previous page)
Institute of Electrical and Electronics Engineers, Washington
Section
Institute of Food Technologies, Washington DC Section
Institute of Industrial Engineers, National Capital Chapter
International Association for Dental Research, American
Section
International Society for the Systems Sciences
International Society of Automation, Baltimore Washington
Section
Instrument Society of America
Marine Technology Society
Maryland Native Plant Society
Mathematical Association of America, Maryland-District of
Columbia-Virginia Section
Medical Society of the District of Columbia
National Capital Area Skeptics
National Capital Astronomers
National Geographic Society
Optical Society of America, National Capital Section
Pest Science Society of America
Philosophical Society of Washington
Society for Experimental Biology and Medicine
Society of American Foresters, National Capital Society
Society of American Military Engineers, Washington DC Post
Society of Manufacturing Engineers, Washington DC Chapter
Society of Mining, Metallurgy, and Exploration, Inc.,
Washington DC Section
Soil and Water Conservation Society, National Capital Chapter
Technology Transfer Society, Washington Area Chapter
Virginia Native Plant Society, Potowmack Chapter
Washington DC Chapter of the Institute for Operations
Research and the Management Sciences (WINFORMS)
Washington Evolutionary Systems Society
Washington History of Science Club
Washington Paint Technology Group
Washington Society of Engineers
Washington Society for the History of Medicine
Washington Statistical Society
World Future Society, National Capital Region Chapter
Richard Hill
Vacant
Neal F. Schmeidler
J. Terrell Hoffeld
Vacant
Vacant
Hank Hegner
Jake Sobin
Vacant
D. S. Joseph
Vacant
Vacant
Jay H. Miller
Vacant
James Cole
Vacant
Eugenie Mielczarek
Vacant
Daina Apple
Vacant
Vacant
E. Lee Bray
Terrell Erickson
Richard Leshuk
Vacant
Russell Wooten
Vacant
Albert G. Gluckman
Vacant
Alvin Reiner
Alain Touwaide
Michael P. Cohen
Jim Honig
Washington Academy of Sciences
Room 113
1200 New York Ave. NW
Washington, DC 20005
NONPROFIT ORG
US POSTAGE PAID
MERRIFIELD VA 22081
PERMIT# 888
Return Postage Guaranteed
w AS
03$^
Volume 101
Number 2
Summer 2015
Journal of the
WASHINGTON
ACADEMY OF SCIENCES
MCZ
LIBRARY
NOV 30 206
HARVARD
UNIVERSITY
Editor’s Comments S. Howard ii
Board of Discipline Editors iii
The Curious Case of Schmidt’s Star T Lipscombe 1
Docosahexaenoic Acid Induces Death in Murine Leukemia Cells by Activating the
Extrinsic Pathway of Apoptosis E E Williams 1 3
A Nineteenth Century Historical Analysis of Game Warden Efforts: Focus on Rabbits
and Hares K. GUcrease 39
Uranus and Neptune Revisited S. Howard 57
Membership Application 77
Instruction to Authors 78
Affiliated Institutions 79
Affiliated Societies and Delegates 80
ISSN 0043-0439 Issued Quarterly at Washington DC
Washington Academy of Sciences
Founded in 1898
Board of Managers
Elected Officers
President
Mina Izadjoo
President Elect
Mike Coble
Treasurer
Ronald Hietala
Secretary
John Kaufhold
Vice President, Administration
Nick Tran
Vice President, Membership
Sue Cross
Vice President, Junior Academy
Vice President, Affiliated Societies
Gene Williams
Members at Large
Paul Arveson
Michael Cohen
Frank Haig, S.J.
Neal Schmeidler
Mary Snieckus
Past President
Terrell Erickson
Affiliated Society Delegates
Shown on back cover
Editor of the Journal
Sethanne Howard
Journal of the Washington Academy of
Sciences (ISSN 0043-0439)
The Journal of the Washington Academy of
Sciences
The Journal is the official organ of the
Academy. It publishes articles on science
policy, the history of science, critical reviews,
original science research, proceedings of
scholarly meetings of its Affiliated Societies, and
other items of interest to its members. It is
published quarterly. The last issue of the year
contains a directory of the current membership
of the Academy.
Subscription Rates
Members, fellows, and life members in good
standing receive the Journal free of charge.
Subscriptions are available on a calendar year
basis, payable in advance. Payment must be
made in US currency at the following rates.
US and Canada $30.00
Other Countries $35.00
Single Copies (when available) $15.00
Claims for Missing Issues
Claims must be received within 65 days of
mailing. Claims will not be allowed if non-
delivery was the result of failure to notify the
Academy of a change of address.
Notification of Change of Address
Address changes should be sent promptly to
the Academy Office. Notification should contain
both old and new addresses and zip codes.
POSTMASTER:
Send address changes to WAS, Rm 113,
1200 New York Ave. NW
Washington, DC 20005
Published by the Washington Academy of
Sciences
email: iournal@washacadsci.org
website: www.washacadsci.org
Academy Office
Washington Academy of Sciences
Room 113
1 200 New York Ave. NW
Washington, DC 20005
Phone:(202) 326-8925
Journal of the
1200 New York Ave
Suite 113
Washington DC
20005
www. wa sh acad scl. org
WASHINGTON ACADEMY OF SCIENCES
Volume 101 Number 2 Summer 2015
Contents
Editor’s Comments S. Howard ii
Board of Discipline Editors iii
The Curious Case of Schmidt’s Star T. Lipscombe 1
Docosahexaenoic Acid Induces Death in Murine Leukemia Cells by Activating the
Extrinsic Pathway of Apoptosis E. E. Williams 13
A Nineteenth Century Historical Analysis of Game Warden Efforts:
Focus on Rabbits and Hares K. Gilcrease 39
Uranus and Neptune Revisited S. Howard 57
Membership Application 77
Instructions to Authors 78
Affiliated Institutions 79
Affiliated Societies and Delegates 80
ISSN 0043-0439 Issued Quarterly at Washington DC
Y1CZ
library
NOV 3 0 2015
harvard
UNIVERSITY
II
Editor’s Comments
In the summer of 2015 we celebrate the diversity of topics
presented in our Journal. The papers in this issue range from stars to
rabbits. Before I briefly describe them, let me say it is an honor to serve
as the new editor of the Journal of the Washington Academy of Sciences.
Our Journal has a long and distinguished history. It is almost unique in its
breadth. Each of you can help continue this history as we go forward.
Please celebrate with me and continue to submit manuscripts on all sorts
of topics: on the many sciences, on technical subjects, engineering, and
mathematics. Expand the field to include the history, sociology, and
psychology of these subjects. I welcome letters to the editor and book
reviews. This is an exciting and challenging task. Join with me as we go
forward.
First up we have a paper by Trevor Lipscombe that reviews the
curious case of Schmidt’s star (first mentioned in 1 891 and then relegated
to the history texts). Trevor resurrects the star to glean possible new
information. To follow it we have a paper by Gene Williams who
discusses fatty acids and cancer. He explains one of the things fish oil is
likely doing for you should you have any cancer cells wandering around
that have not been killed off by the immune system. Third we have Kelsey
Gilcrease writing about the efforts of 19th century game wardens (chasing
those rabbits) in New Jersey and Massachusetts. To complete this issue I
include a history paper on how the rotational periods (the lengths of their
day) of Uranus and Neptune were determined before the space mission
Voyager traveled by them.
In the 2007 Spring issue, Vol 93, we published an article by Y.
Said (then at George Mason University): “On the Eras in the History of
Statistics and Data Analysis”. We have since retracted this article because
of suggested controversy over its uniqueness.
Sethanne Howard
Editor
Washington Academy of Sciences
Ill
Journal of the Washington Academy of Sciences
Editor Sethanne Howard
sethanneh@msn.com
Board of Discipline Editors
The Journal of the Washington Academy of Sciences has a 12-member
Board of Discipline Editors representing many scientific and technical
fields. The members of the Board of Discipline Editors are affiliated with
a variety of scientific institutions in the Washington area and beyond —
government agencies such as the National Institute of Standards and
Technology (NIST); universities such as Georgetown; and professional
associations such as the Institute of Electrical and Electronics Engineers
(IEEE).
Anthropology Emanuela Appetiti
Astronomy Sethanne Howard
Biology/Biophysics Eugenie Mielczarek
Botany Mark Holland
Chemistry Deana Jaber
eappetiti@hotmail.com
sethanneh@msn.com
mielczar@physics.gmu.edu
maholland@salisbury.edu
diaber@mai~vmount.edu
Environmental Natural
Sciences Terrell Erickson
Health Robin Stombler
History of Medicine Alain Touwaide
Operations Research Michael Katehakis
Physics Katharine Gebbie
Science Education Jim Egenrieder
Systems Science Elizabeth Corona
terrell.ericksonl@wdc.nsda.gov
rstombler@aubumstrat.com
atouwaide@hotmail.com
mnk@rci.rutgers.edu
katharine.gebbie@nist.gov
i im@deepwater.org
elizabethcorona@gmail.com
Summer 2015
Washington Academy of Sciences
1
The Curious Case of Schmidt’s Star
Trevor Lipscombe
Catholic University of America Press, Washington, DC.
Abstract
This article discusses the internal structure of a type of star first
proposed by August Schmidt in 1891, one that causes any light to enter
it to move in a circle. An exact analytical solution of the equation of
hydrostatic equilibrium is thus obtained. The solution is physically
realistic, in the sense that the central density, central pressure, and total
mass are all finite, while both density and pressure drop to zero at the
outer radius of the star. In the core of the star, the pressure depends only
weakly on density. The outer layers of the star can be well-
approximated as isothermal. Schmidt’s star, then, is a physical system
of historical, pedagogical, and mathematical interest.
Introduction
In the late 1800s astrophysicists faced a conundrum. The age of the
Earth had been reliably determined and thus the minimum age of the Sun
was also known; but given that age, and the known laws of physics, stars
should have burnt out long before.1 We now know that the stars shine
because of nuclear processes that take place in their core, processes
unknown in the nineteenth century.
In 1891 a German scientist, August Schmidt (1840-1929),
proposed a radical solution to resolve the paradox. What if stars didn't
shine but were, in a sense, mirages? That is to say, suppose stars acted as
giant lenses, with a refractive index that varies as a function of radius and
that causes any ray of light to enter it to move with a circular motion,
thereby never leaving the star? This would remove any need for a
mechanism by which stars had to bum fuel to generate energy, and so
resolve the conundrum.
Schmidt proposed a model in which the Sun’s outer surface was
such an optical illusion2. This idea caught the attention of “a student of
astronomy” E.J. Wilczynski, who wrote the first English-language article
on Schmidt’s theory in 1895. It appeared in the first-ever volume of the
Astrophysical Journal3 (ApJ), followed only a few pages later by a note
Summer 2015
2
from James Edward Keeler, the co-founder and co-editor of the ApJ, who
commented:
“...The theory is apt to be more favorably regarded by mathematicians
than by observers' ,4
a sentiment echoed by George Ellery Hale— the other co-founder and co-
editor of the ApJ — who, in the very next volume, wrote :
“As a theoretical discussion the theory is interesting and valuable, but few
observers of the Sun wall consider it capable of accounting for the varying
phenomena encountered in their investigations"5 .
Here, though, we follow Michael Faraday’s dictum that “Nothing
is too wonderful to be true, if it be consistent with the laws of nature" and
investigate Schmidt’s theory, to see whether such an astrophysical object
could actually exist. Heretofore, studies have only paid attention to the
optical properties of Schmidt’s star. (The exact nature of the outer visible
layer of the Sun, which was in large part what Schmidt dealt with, still
generates controversy.6)
In this article, we determine the physical properties of Schmidt’s
star, which appears not to have been done before. The basic mechanism
proposed by Schmidt leads to an exact solution for the equation of
hydrostatic equilibrium, which governs self-gravitating stationary spheres
of fluids.
Density Determination
In Waves and Grains , Mark Silverman analyzes the optical
properties of Schmidt’s star7. If a spherical lens has a refractive index n(r)
and possesses spherical symmetry, then Silverman shows that a point on
a light ray r is given by:
dr
He
(1)
where R is the radius of the star. For the light rays to move in a circular
path, we require that r = constant, and the refractive index in the medium
must therefore vary as:
n{r ) = —■ (2)
r
Washington Academy of Sciences
3
Note that when r — R we have n - 1, which is the refractive index of the
vacuum.
The refractive index in a medium depends, among other things, on
the density of that medium, which is the basic mechanism behind the
formation of mirages. The Clausius-Mosotti (or Lorentz-Lorenz) relation
can be used to relate the refractive index of a substance to its density8:
rr - 1
n2+ 2
Kp
(3)
where K is a constant that depends on the particular gas of which the star
consists. Studies of the Lorentz-Lorenz relation for gaseous and liquid
hydrogen show that K remains approximately constant (-1.03 cm”3 / g)
for a broad range of temperatures (15-298 Kelvin) and pressures (9-200
atm)9, though it may well not hold at the high densities typically found at
the centers of stars. Substituting in from Eq. (3) above:
R2-r 2
R2+2r2
- Kp.
(4)
When r = R, the density falls to zero, as it should at the outer radius of
the star. Note also that when r = 0, Kp{ 0) = 1, so that K is the inverse of
the central density pc. Thus, using the scaled dimensionless radius
x = r / R, we can write:
P = Pc
^1-x2^
1 +2x"
(5)
As a consequence of knowing how the density varies within this
astrophysical object, we can calculate its mass, M:
i
M = 47rpcR2\
x 2 (l-x2 )
1 + 2x"
<Ix
(6)
which integrates to:
M - 4np( R
12 8
(7)
and hence numerically:
Summer 2015
4
M ~ 4k (0.07 67) p( R3 .
If the average density is /?, then by definition:
(8)
4 ttR3 _
M - — - — p.
(9)
which means that the average density of Schmidt’s star is related to the
central density by:
p = 0.2301/?,
c-
(10)
Pressure Determination
The equation of hydrostatic equilibrium of a static, self-gravitating
sphere of fluid of density p(r) and pressure P(r) is10:
1 d
r2 dr
' r_dp
yP dr ,
= 4 7iGp{r)
(11)
or, in terms of the dimensionless radius,
1 d
x2 dx
v rf/y
v p dx
= -4/rGR2 p(x).
(12)
Thus, by means of Eq. (5),
Vdfp
v p dx
d
dx
= -AkGR2 pcx~
f
1 -x
2 \
1 + 2x2
(13)
Integrating:
— — ^- = -4 7rGR~ p — -4x3 +18.X-9V2 tan 1 V2x + const. (14)
p(x) dx 24 L 3
Given that the left-hand side of the equation is zero at the center of the
star, the constant must be zero. Hence we have:
dP 7rGR2p2
dx
4x +
18 9V2
tan
V2;
M-x2 '
1 + 2x
(15)
This can be integrated term-by-term analytically once again to give:
Washington Academy of Sciences
5
m
_ tKjR2pI
j-2x2 +31n(2x2 + l)-l}-18*j-— — h— + lnx
4 2
>+-
9V2
{Vlx
x2 -51n(2x2 +l) + 61nx + -4x4+6x2-6
Eq. (16) simplifies to:
tan
6x
+ const.
(16)
1 7 2 9 4 27
7x — x
2 2 2
+ 9V2
ln(2x2 +l)
(17)
+const.
At the outer radius of the star, the pressure drops to zero and thus P(x = 1 )
= 0. This boundary condition allows for the calculation of the integration
constant, since we must have:
P{x = 1) = 0 =
KGRjpl
1 9 27
7-- In 3
2 2 2
+ const.
(18)
Hence:
7cGR2 p2
const = —
1 9 27, 0
— + 7 H 1 In 3
2 2 2
(19)
The complete solution for the pressure distribution within Schmidt's star
is thus:
. 7rGR 2 p2.
P(x) =
^23 + 27 In 3^ „ , 9 4
- lx" — x
2
J
27
f 1 A
In ^2x2 +l) + 9v2 x — tan”1 V2x
v X )
(20)
Note that:
f
lim
x— >0
tan 1 >[2x
x
(21)
so that the central pressure is indeed finite and has the value:
Summer 2015
6
p _ ”GR2Pc
rc ~
^ 27 In 3 — 1 3 ^
v 2 ,
(22)
Hence:
Pc
m=-
^23 + 27 ln3N
2 ,
|x4-yta(2xJ+l)+9\/2
7 271n3— 13V
l 2 J
(23)
A graph of the pressure and density within the star, as a function of stellar
radius, is shown in Figure 1.
Fig.l Density and Pressure as a Function of Radius
Density Pressure from Eq. (20)
The pressure and density of the star are such that a good
approximation for their relation is1 1 :
P = \.0\\\2PC
1 - exp
^-4,1 A)
V Pc y
(24)
The exact relation of pressure with density [from Eqs. (5) and (20)] is
compared with the approximate relationship of Eq. (24) in Figure 2.
Washington Academy of Sciences
7
Hence, by starting with a simple requirement — light rays travel in
circles within the astrophysical object — we can determine that such a star
has a finite central density and pressure and determine the variation of the
pressure and density as a function of the radius.
Comparison with Poiytropic Models
The usual approach to stellar astrophysics is to explore the
polytrope equation12. That is to say, one seeks solutions of the form:
P = KpMln (25)
in the equation for hydrostatic equilibrium. This model generates the
Lane-Emden equation, named after Jonathan Homer-Tane, an
astrophysicist who spent many years in Washington DC, and Swiss
astrophysicist Robert Emden. Here n is not the refractive index, but the
so-called polytropic index of the star.
It is complicated to compare Schmidt’s star with standard
polytrope solutions. For example, as seen from Eq. (10), the central
density is related to the average density by:
pc = 4.348/? (26)
Summer 2015
8
which, by use of the Polytrope Tool13, is equivalent to the central pressure
of a polytrope whose index is n — 1.31. The standard model for the Sun
(the Eddington model, for which n = 3) has the value pc =54.18 p, just
over a factor of ten larger14.
The central pressure in Schmidt’s star is given by numerically
evaluating Eq. (22):
P.. =8.331 26 P' . (27)
6
The mass, though, is given in Eq. (8), as:
M ~ 4xpcRi (0.0767) (28)
and so by substituting in for the central density, we obtain:
8.33126;rGfl2 ^ w v
P,
c
M
4/rR3 (0.0767)
(29)
or:
R. * 4.7
c
GM 2
R4
(30)
Again, by means of the Polytrope Tool, this is an expression equivalent to
the central pressure of a polytrope of index n = 2.595, almost double the
index obtained from consideration of the central density. The Eddington
model for the Sun has the numerical factor 1 1 .05 rather than 4.7, a central
pressure some 2.35 times higher than the Schmidt star.1''
For further comparison, note that for a polytrope:
P_
T
(jl|
V Pc J
1+1/ n
Consequently:
P_p£ = fP_\n
Pc P 'Pc'
(31)
(32)
Washington Academy of Sciences
9
We can thereby define an effective pointwise polytropic index n by:
n =
(33)
A curve of n as a function of radius is shown in Figure 3.
Fig. 3 Effective polytropic index as function of radius
Index, n
The best fits have n = —1.22 from x = 0 to 0.5, half way out
through the star. This is a negative polytrope of varying index.
Approaching x = 1, the index becomes large and negative, so that
P ~ const p , which is the equation describing an isothermal outer layer
(infinite polytropic index) to Schmidt’s star.
Discussion
Negative-index polytropes were first discussed by Eddington in
1931 16. In the same paper, he used the phrase “Incomplete poly tropes” to
describe a structure similar to Schmidt’s star, wherein the inner core might
best be modeled by one value of the polytropic index and the outer layers
by another. Viala and Horedt17 showed that astrophysical objects with
negative polytropic indexes are good models for, among other things,
interstellar clouds. In addition, they showed that sufficiently negative
Summer 2015
10
indexes ( n < 1), can be stable. Chaplygin gases, which have negative
indexes, are currently of great interest in cosmology, as they are
candidates for dark matter and can form stable gravitational structures
(both in classical Newtonian and general relativistic gravity)18.
In Schmidt's star, both the pressure and the density decrease when
moving radially outwards. In the outer layers, the pressure varies almost
linearly with density, which suggest an isothermal envelope for the star.
However, in the stellar core it is similar to a negative-index polytrope of
index n < H, in that the temperature must increase radially outwards. This
creates a significant problem for Schmidt's star, in that to be physically
realistic, the model must represent an astrophysical object whose core is
being heated externally, either by particles or by radiation, in a spherically
symmetric manner, but whose outer layers are isothermal.
Keeler and Hale's original criticisms of Schmidt’s proposal were
that it was of importance only mathematically. Regrettably that may
indeed be the case. Schmidt’s star, though, remains of interest. Such
interest is not just historic; Schmidt’s star also is of pedagogical value19.
Undergraduate physics students, as an exercise in physical modeling,
could be presented with the Schmidt-Silverman equation for the refractive
index of Schmidt’s star and then asked to solve for the pressure and
density of this object. This requires knowledge of various disciplines
within physics. They could also be asked to comment on whether such an
object could indeed exist, which requires them to recognize that the
temperature in an astrophysical object should, to be realistic, fall off with
increasing radius.
Conclusions
In this paper, we have explored the structure of an astrophysical
object whose physics has not previously been determined completely. By
requiring such a star to act as a graded refractive index lens that causes all
light entering in to it to move in a circular path, we have been able to
determine the density of the star and its pressure. Such a star has a
pointwise negative polytropic index, but its pressure and density both
decrease as the radius increases, and the value of the index is such that
Schmidt’s star is likely to be stable. While the study was motivated by
Schmidt’s suggestion in 1891, the density profile and pressure profile here
represent an exact solution of the equation of hydrostatic equilibrium.
Washington Academy of Sciences
II
whether one gives credence to Schmidt’s belief or not. This is one of the
few physically motivated stellar models, other than polytropes, that is not
singular at the origin nor infinite in extent. While the temperature profile
makes Schmidt's star likely to be physically unrealistic, it remains of
historical, mathematical, and pedagogical value.
This paper is dedicated to Kelsey Schmidt and Tom LaCour on the
occasion of their marriage.
REFERENCES
1 Frank D. Stacey ‘Kelvin’s Age of the Earth paradox revisited,” Journal of
Geophysical Research 105(B6), pp 13155-13158 (2000).
2 August Schmidt “Die Strahlenbrechung auf der Sonne: ein geometrisches Beitrag zur
Sonnenphysik” (Stuttgart: Metzlerscher Verlag, 1891).
3 Ernest J Wilczynski “Schmidt’s Theory of the Sun”, Astrophysical Journal vol. 1, pp
112-126 (1895).
4 James E Keeler “Schmidt’s Theory of the Sun”, Astrophysical Journal vol. 1, pp 178-
179 (1895).
5 George E Hale “Notes on Schmidt’s Theory of the Sun”, Astrophysical Journal vol.
2, pp 69-74 (1895).
6 See, for example, Pierre-Marie Robitaille “Commentary on the Radius of the Sun:
Optical Illusion or Manifestation of a Real Surface”, Progress in Physics Vol. 2,
L5-L6 (2013).
7 Mark P. Silverman Waves and Grains (Princeton N.J.: Princeton University Press,
1998), pp 27-30.
8 See, for example, Charles Kittel Introduction to Solid State Physics (8,h edition) (New
York: John Wiley & Sons, 1990).
9 Dwain E. Diller “Refractive Index of Gaseous and Liquid Hydrogen” J. Chem Phys.
49(7) 3096-3105 (1968).
10 See, for example, S. Chandrasekhar An Introduction to the Study of Stellar Structure
(Chicago: University of Chicago Press, 1939), page 63, equation (6).
11 By inspection, a trial solution is P — const[ 1 — exp(— ap)]. The requirement that
P = 1 when p ~ 1 determines the constant in terms of a. At low densities, we
have P = [a/[ 1 — exp(— a)]p and so use of the data at low densities in a linear
regression estimator (such as LINEST in Microsoft Excel) leads to the best
numerical fit.
12 S. Chandrasekhar An Introduction to the Study of Stellar Structure (Chicago:
University of Chicago Press, 1 939), pp. 84-182.
Summer 2015
12
L' http://www.webnucleo.Org/home/online_tools/polytrope/0.8/
14 S. Chandrasekhar An Introduction to the Study of Stellar Structure (Chicago:
University of Chicago Press, 1939), equation 56, chapter 6, page 230.
15 S. Chandrasekhar An Introduction to the Study of Stellar Structure (Chicago:
University of Chicago Press, 1939), chapter 6, page 230, equation 57.
16 Arthur S. Eddington “A Theorem Concerning Incomplete Polytropes” Mon. Not.
Roy. Ast. Soc. 91 pp 440-444 (1931).
17 Yves P. Viala and Georg P. Horedt “Polytropic Sheets, Cylinders, and Spheres with
Negative Index”. Astron. & Astrophys. 33 pp 195-202 (1974).
18 *i
Trevor C. Lipscombe “Self-gravitating clouds of generalized Chaplygin and anti-
Chaplygin gases,” Physica Scripta 83(3) ID = 035901 (201 1).
19 Another example of historically motivated physics of pedagogical value might be the
modeling of Newton’ s-bucket experiment in Carl E. Mungan and Trevor C.
Lipscombe “Newton’s Rotating Water Bucket: A Simple Model,” Journal of the
Washington Academy of Sciences 99(2), pp 15-24 (2013).
Bio
Trevor Lipscombe is the director of the Catholic University of America
Press. He holds a doctorate in theoretical physics from Oxford, is a Fellow
of the Royal Astronomical Society, and tries to do theoretical physics in
his spare time. He is the author of “The Physics of Rugby’' (Nottingham
University Press, 2009); coauthor, with Alice Calaprice, of “Albert
Einstein: A Biography” (Greenwood, 2005); and editor of a critical
edition of Blessed John Henry Newman's novel “Loss and Gain: The
Story of a Convert” (Ignatius Press, 2012).
Washington Academy of Sciences
13
Docosahexaenoic Acid Induces Death in Murine
Leukemia Cells by Activating the Extrinsic Pathway
of Apoptosis.
E. Eugene Williams
Salisbury University
Abstract
Docosahexaenoic acid (DHA) is a unique fatty acid that is found
predominantly in the phospholipids of cell membranes. It has wide-
ranging therapeutic effects that are broadly appreciated but poorly
understood. Its principal location in the membranes of cells suggests that
these myriad effects are manifest there. When cultured in DHA-enriched
medium, cells of the murine leukemia cell line T27A took up the fatty
acid and incorporated it into cellular phospholipids, particularly those of
the plasma membrane. Culture in DHA-enriched media also caused
significant dose-dependent cell death accompanied by increased plasma
membrane bleb formation. Cysteine-dependent aspartate-directed
proteases (caspases)-3 -8 and -9 were also activated, establishing
apoptosis as the mechanism of DHA-induced cell death. Inhibition of any
one of these caspases rescued the cells from apoptotic death. Caspase
inhibition experiments identified T27A cells as belonging to the type II
group of apoptotic cells and showed that apoptosis was initiated via the
extrinsic pathway. Together these and previous data support the
hypothesis that DHA causes cell death in leukemic cells by inducing
alterations in the structure of lipid rafts that lead to the ligand-independent
activation of death receptors and apoptosis.
Introduction
Docosahexaenoic acid (DHA, 22:6n-3) is a unique fatty acid that is
found in the cells of a wide range of organisms from bacteria to humans.
It is the longest and most unsaturated of the commonly occurring n-3
(omega-3, co-3) fatty acids (Salem et al. 1986). DHA has diverse
therapeutic properties that are acclaimed in both the scientific and lay
communities (Stillwell and Wassail 2003; Siddiqui et al. 2004; Chapkin
et al. 2009). A remarkable number of conditions and diseases have been
demonstrated to be prevented, mitigated, counteracted or improved by
DHA. These include maladies as disparate as cancer, heart disease, cystic
fibrosis, diabetes, immune function and even psychiatric disorders
(Stillwell and Wassail 2003; Siddiqui et al. 2004; Calder 2012;
Mischoulon and Freeman 2013). While the relationship between DHA
Summer 2015
14
and improved health is widely appreciated, the basic molecular
mechanism underlying this relationship remains unclear. As noted by
Stillwell (2008), the assortment of seemingly unrelated biochemical and
physiological processes underlying the diseases and conditions that are
influenced by DHA suggests that this fatty acid influences a fundamental
cellular function or property.
DHA has been shown to have powerful anti-cancer effects in
animals and cultured tumor cells (Siddiqui et al. 2004). For example, it is
effective at reducing the accumulation of leukemic cells in vitro and in
slowing the rate of progression of leukemia in animals ( e.g . Jenski et al.
1993; Jenski et al. 1995; Zerouga et al. 1996). DHA has been shown to
induce cell death in human and mouse leukemia cells in a dose dependent
manner (Kafrawy et al. 1998; Yamagami et al. 2009) and it has been
suggested the anti-leukemia properties of DHA are in general founded on
the ability of DHA to induce cell death in tumor cells (Serini et al. 2009).
Despite continuing efforts, it is currently unclear precisely how
DHA triggers cell death. DHA can be converted into reactive oxygen
species that can influence cell survival (Siddiqui et al. 2008), and into
powerful anti-inflammatory and pro-resolving mediators (resolvins,
protectins and maresins) that can influence cell survival and disease
etiology (Serhan et al. 2014; Colas et al. 2014; Dalli et al. 2015). DHA
can also affect gene expression (Berger et al. 2006), the acylation patterns
of membrane proteins (Webb et al. 2000), and the function of enzymes
and ion channels (Matta et al. 2007). However a large and growing body
of evidence indicates that DHA induces cell death only after it has become
incorporated into membrane phospholipids and that the initial triggering
event in cell death is a membrane-based phenomenon (Stillwell and
Wassail 2003; Stillwell et al. 2005; Calder 2012).
There is substantial physiological, biochemical, biophysical, and
morphological evidence that DHA-containing phospholipids change the
structure of cell membranes (Mitchell et al. 2003; Niu and Mitchell 2005;
Chapkin et al. 2008; Shaikh 2010; Rockett et al. 2012; Teague et al. 2013;
Pinot et al. 2014). Indeed, whether provided as a dietary component to an
individual organism (Lien 2009) or as a component of the incubation
medium of cultured cells (Zerouga et al. 1996; Williams et al. 1998;
Williams et al. 1999), DHA is taken up by cells and incorporated into the
Washington Academy of Sciences
15
phospholipids of membranes. The plasma membrane in particular appears
to be a primary location of action for the tumor cell killing properties of
DHA (Jenski et al. 1993; Pascale et al. 1993; Williams et al. 1998;
Williams et al. 1999). Of particular interest in this regard is the influence
of DHA-containing phospholipids on the membrane microdomain
structures known as lipid rafts. Lipid rafts serve as platforms for the
regulation of cell processes and represent a selective cellular compartment
that can co-localize and modulate the activities of enzymes, receptors and
other proteins (Simons and Ikonen 1997; Lingwood and Simons 2010).
There is evidence that DHA-containing phospholipids induce cell death
by altering the structure or organization of lipid rafts, and that this
influence on membrane structure is the first and most important step in
DHA-induced cell death (Stillwell et al. 2005; Schley et al. 2007; Chapkin
et al. 2008).
Other evidence strongly suggests that DHA causes cell death in
tumor cells by the induction of apoptosis (Blanckaert et al. 2010; Kang et
al. 2010). There are two distinct activation pathways for apoptosis. The
extrinsic pathway involves plasma membrane-associated death receptors
and a cysteine-dependent aspartate-directed protease, caspase-8. The
intrinsic pathway involves the release of cytochrome c from mitochondria
and the activation of caspase-9. These two initiating events then cause the
activation of downstream effector caspases including caspase-3 which in
turn cleaves a series of intercellular substrates to continue the apoptotic
cascade. Lipid rafts are importantly involved in the extrinsic apoptotic
pathway as the death receptors, a subset of the tumor necrosis factor
receptor superfamily, are among those receptors regulated by lipid rafts
(Gajate et al. 2009; Lang et al. 2012).
Thus, there is evidence that DHA causes the death of many types
of tumor cells, that the cause of cell death in many of these instances is
the induction of apoptosis, that DHA alters the structure of lipid rafts, and
that lipid rafts regulate the receptors involved in initiating the extrinsic
pathway of apoptosis. This study attempts to connect these links by testing
the hypothesis that DHA causes cell death in leukemia cells by
specifically triggering the extrinsic pathway of apoptosis. We show that
DHA is selectively incorporated into the plasma membrane of murine
leukemia (T27A) cells. We use both morphological and biochemical
Summer 2015
16
means to demonstrate that DHA induces apoptosis in these cells. By
monitoring and manipulating the activities of caspases -8, -9 and -3, we
further show that all three caspases are activated by DHA and that the
inhibition of any one of them rescues T27A cells from DHA-induced
apoptosis. Together these and previous data support the hypothesis that
DHA causes cell death by inducing alterations in the structure of lipid rafts
that lead to the ligand-independent activation of death receptors and
apoptosis.
Methods
Materials
T27A murine leukemia cells were obtained from American Type
Culture Collection (Manassas, Va). Fatty acids and fatty acid methyl ester
(FAME) reference standards were purchased from Nu-Chek-Prep
(Elysian, MN). RPMI-1640 culture medium supplemented with 2 mM
glutamine, 25 mM HEPES, 50 pg/mL streptomycin and 100 units/mL
penicillin, was from Cambrex Bio Science (Walkersville, MD). Bovine
calf serum was from Hyclone (Logan, UT). Irreversible, cell-permeable
inhibitors of caspases -3 (Z-D[0-Me]E[0-Me]VD[0-Me]-FMK), -8 (Z-
IE[0-Me]TD[0-Me]-FMK), and -9 (Z-LE[0-Me]HD[0-Me]-FMK)
were from Calbiochem (EMD Biosciences, Inc., La Jolla, CA). The
colorimetric assay kits for measuring the activities of caspases -3, -8 and
-9 were from BioVision (Mountain View, CA.). Staurosporin, SiCE
(“Celite”), and dimethyl sulfoxide (DMSO) were from Sigma Chemical
Co. (St. Louis, MO). All other chemicals were from Sigma or Thermo
Fisher Scientific (Waltham, MA).
Cell culture
Except where noted, T27A cells were cultured in RPMI-1640
medium supplemented as described above and with 10% (vol/vol) bovine
calf serum in 25 cm2 culture flasks maintained at 37°C under an
atmosphere of 5% CO2 in humidified air. As noted previously (Zerouga
et al. 1996; Williams et al. 1998; Williams et al. 1999), under these
conditions cultures doubled every 12 to 15 hours. Cell viability was
monitored by trypan blue exclusion (0.04% in phosphate buffered saline
[PBS, 0.154 M NaCl, 0.016 M NaH2P04, pH 7.2]).
Washington Academy of Sciences
17
Supplementation of culture media with fatty acids
DHA and oleic acid (OA, 18:1 n-9) were added to RPMI culture
medium using the methods of Spector and Hoak (1969) exactly as
described by Williams et al. (1998). The fatty acid was dissolved in
hexane and transferred to an Erlenmeyer flask containing SiCU Ten g of
SiCb were used per mmol of fatty acid. The hexane was removed
completely by a gentle stream of N2 before the dry mixture was transferred
to a solution of fatty acid free bovine serum albumin (1% fatty acid free
BSA in RPMI supplemented as above, but excluding serum). After
stirring for 30 min in the dark, the RPMI/fatty acid mixture was
centrifuged for 30 min at 600 grav to remove the Si02 and the medium
was sterilized by filtration (0.22 pm). Bovine calf serum was added to
10% (vol/vol) of the total just before use. Calf serum contributes a small
amount of fatty acids to the final culture medium, but less than 1 % of that
is DHA (Williams et al. 1998). Unless noted otherwise, cells were
incubated in fatty acid-enriched medium for 3 days (68-76 hours). Under
these conditions T27A cells take up considerable DHA, and at DHA
concentrations below 0.61 mM they remain >90% viable (Williams et al.
1998; Williams et al. 1999; and see below).
Assay of caspase activity and caspase inhibition
The activities of caspases -3, -8, and -9 were measured
spectrophotometrically in 90-well plates. For each assay, T27A cells were
cultured in RPMI medium containing no additions, 1.3 pM staurosporin,
or 0.61 mM DHA. After 16 h of culture, cells from each flask were
harvested by low-speed centrifugation. Cell viability (always greater than
90% in control cells) was assessed by trypan blue exclusion and cell
density was determined by duplicate counts on a hemacytometer. For each
treatment, 3 x 106 cells were treated with 50 pF of lysis buffer according
to the manufacturer’s instructions. After centrifugation, 30 pF of cell
lysate were mixed with 20 pF of caspase assay medium in a well of the
plate, mixed, and allowed to incubate at 37°C for 1 hour before the
absorbance was read at 405 nm. Background values were subtracted from
all absorbances and all treatment values were expressed as percentage of
the control. For caspase inhibition experiments, cells were exposed to 10
pM inhibitor in DMSO (0.1% final concentration) for 30 min before
exposure to control or fatty acid-enriched medium.
Summer 2015
18
The linearity of the caspase assays was confirmed using p-
nitroanaline as a standard. Regression analyses of the resulting standard
curves yielded lines with n >0.990. Staurosporin was used as a positive
control and only those assays that showed staurosporin-induced caspase
activity were analyzed further.
Isolation of plasma membranes
After 48 h of culture in either normal (control) medium or medium
enriched with 0.3 mM DHA, cell cultures were disrupted by sonication
and the resulting homogenate was fractionated by the centrifugation
protocol of Kaduce et al. (1977) using the buffers of Molnar et al. (1969)
as described by Williams et al. (1998; 1999). Briefly, T27A cells were
collected by centrifugation (500 grav for 15 min), resuspended in 0.25 M
sucrose buffer (0.25 M sucrose, 40 nrM NaCl, 100 mM KC1, 5 mM
MgSO-t. 7 HrO, 20 mM Trizma base, pH 7.2 with HC1), and disrupted (on
ice) by sonication for 2 x 35 sec using a tip-type sonicator (Fisher
Scientific Model 500, 35 seconds, pulse on 1 sec, pulse off 1.5 sec). The
cell homogenate was centrifuged at 27 kgrav for 10 min to remove
undisrupted cells and cellular debris and the supernatant over the resulting
pellet was spun for 1 hour at 105 kgrav to produce a mixed membrane
pellet. The mixed membrane pellet was layered onto a pad of 1.1 M
sucrose (remaining composition as above) and spun at 107 kgrav for 16
hours. The white interfacial material was collected and washed twice in
excess PBS. The resulting membrane represents a better than 8-fold
purification of plasma membrane over the mixed membrane fraction
(Kaduce et al. 1977) and has been used in previous studies to determine
the effects of DHA on membrane structure and composition in T27A cells
(Williams et al. 1998; Williams et al. 1999).
Lipid extraction and gas chromatography of membrane fatty acids
Total lipids where extracted from whole cell preparations and
from isolated plasma membranes using CHCI3/CH3OH (Bligh and Dyer
1959) and concentrated under a stream of dry N2 gas. Phospholipids
separated from neutral lipids ( e.g ., triacylglycerols) by silicic acid
chromatography (Wren 1960; Williams and Somero 1996) were
transesterified into FAMEs using methanolic sodium methoxide (Eder et
al. 1992). FAMEs were resolved using a 0.25 mm x 30 nr HP-23 cis/trans
Washington Academy of Sciences
19
FAME column in a Hewlett-Packard 6890 gas chromatograph. The
instrument was programmed to produce a temperature ramp from 1 80°C
to 240° at 2°C/min starting 2 minutes after sample injection. Peaks
corresponding to individual FAMEs were identified by comparison ot
retention times to those of authentic standards. Peak areas were calculated
using Hewlett-Packard’s ChemStation software.
Statistics
Statistical analyses were carried out using version 2.15.3 of R (R
Development Core Team, 2008; http://www.r-project.org/).
Probabilities < 0.05 were considered significant (and labeled *). Percent
data were arcsine transformed (sin'Wproportion) before statistical
analyses as recommended (Sokal and Rohlf 1981). The normality of
distribution of each data set was assessed using the Shapiro-Wilk test. The
homogeneity of variances among data sets was tested using Fligner-
Killeen test as it has been shown to be least sensitive to departures from
normality (Conover et al. 1981). The slopes of regression lines were
compared to each other and to slope = 0 using the linear model function
of R. Group means were compared using one-way analysis of variance
(ANOVA) followed by Tukey's HSD mean separation test, or where
appropriate, the Kruskal-Wallis test followed by Wilcoxon rank sum tests.
Results
When cells of the murine leukemia line T27A were cultured in
media supplemented with DHA, they took up the fatty acid and
incorporated it into cellular phospholipids (Table 1). In phospholipids
isolated from whole cells, DHA levels were 25 times that found in control
cells. The increased proportion of DHA was associated with a large
reduction in the proportions of stearic acid (18:0) and the n-6 isomer of
18:3. Proportions of palmitic acid (16:0) and oleic acid (18:1) increased.
By contrast, DHA incorporation into phospholipids of the plasma
membrane represented an 1 8-fold increase over that of control cells and
resulted in a final proportion of DHA almost twice that observed in
phospholipids isolated from whole cells. In the plasma membrane, DHA
largely displaced oleic acid (1 8:1), as well as arachidonic acid (20:4) and
other long chain polyunsaturated fatty acids. The DHA-induced alteration
of membrane lipid composition of both whole cells and plasma membrane
Summer 2015
20
is reflected in the near inversion of the n-6/n-3 ratios (Table 1) after
treatment with DHA.
The dramatic accumulation of DHA in phospholipids of the
plasma membrane of T27A cells can also be clearly seen when comparing
the ratios of palmitate to stearate (16:0/18:0) and of DHA to stearate
(22:6/1 8:0) in phospholipids extracted from plasma membrane and whole
cells cultured in control versus DHA-enriched media (Figure 1). The
results shown in Table 1 and Figure 1 closely mirror previously reported
observations on the effects of DHA on the lipid composition of plasma
membranes isolated from these cells (Zerouga et al. 1996; Williams et al.
1998; Williams et al. 1999) and indicate that the experiments presented
here both compliment and expand those earlier works.
Table 1. The distribution of phospholipid fatty acids, as percent of total
fatty acids, extracted from whole cells and from isolated plasma
membranes after 48 h of culture in control or DHA-enriched (0.3 rnM)
medium. Minor fatty acids, i.e. those comprising less than 1% of the total,
are excluded from the analysis. The data represent the means of two
independent experiments.
Washington Academy of Sciences
21
Figure 1. The ratios of mean values of palmitate to stearate (16:0/18:0) and of DHA to
stearate (22:6/18:0) in phospholipids extracted from whole cells or plasma membrane
(PM) after the cells had been cultured for 48 h in control medium or in medium
containing 0.3 mM DHA.
Figure 2. The density of T27A cells three days after exposure to the indicated
concentrations of fatty acid and expressed as a percentage of control. Squares, OA;
circles, DHA. Linear modeling revealed that the slope of the regression of the OA
response is not significantly different from zero. The slope of the regression of the DHA
response is highly significantly different from both slope = 0 and the OA response (p <
0.001 in both cases). Each point represents the mean ± 1 standard error of the mean from
n = 7-14 (DHA) or n = 3-6 (OA) independent determinations of different cultures.
Summer 2015
22
Culture in DHA-enriched medium caused a significant reduction
in the rate of leukemic cell proliferation. Figure 2 shows a DHA-dose
dependent reduction in cell density compared to control cultures and to
cultures similarly exposed to OA. The proportion of viable cells in the
DHA-enriched cultures also fell significantly, while the viability of cells
in cultures exposed to OA remained indistinguishable from that of the
controls (Figure 3). Together these data show that DHA caused significant
cell death over a three day exposure to concentrations of DHA in the
culture medium from 0.3 to 0.9 mM.
Phase contrast microscopy revealed that unlike control cells or
cells cultured in OA-enriched medium, cells cultured in DHA-enriched
medium were irregularly shaped and exhibited conspicuously higher
internal complexity including extensive cytoplasmic vacuolization. In
addition, the external surfaces of control cells and of cells cultured in OA-
enriched medium were even and regular, whereas the surfaces of cells
cultured in DHA-enriched medium were uneven and displayed numerous
exvaginations of the plasma membrane (commonly referred to as “blebs”;
e.g. Charras 2008). Figure 4 shows that the percentage of T27A cells
exhibiting blebs increased steadily with DHA dose until at the highest
doses tested these structures appeared on nearly 75% of all cells present
in the culture.
Culture of T27A cells for 16 h in a medium containing 0.61 mM
DHA resulted in a significant elevation of the activities of caspases-3, -8,
and -9 (Figure 5). When cell cultures were individually treated with 10
pM of an inhibitor specific for each of these caspases for 30 min prior to
culture in DHA-enriched medium they did not undergo cell death and cell
densities were similar to those of control cultures (Figure 6).
Washington Academy of Sciences
23
Figure 3. The viability of T27A cells as assessed by trypan blue exclusion three days
after exposure to the indicated concentrations of fatty acid. The slope of the regression
of the OA response is not significantly different from zero and that of the regression of
the DHA response is highly significantly different from both slope = 0 and the OA
response (p < 0.001 in both cases). Squares, OA; circles, DHA. Each point represents the
mean ± 1 standard error of the mean from n = 3 independent determinations of different
cultures.
0.0 0.2 0.4 0.6 0.8
[Fatty Acid] (mM)
Figure 4. The percentage of T27A cells exhibiting plasma membrane exvaginations
(“blebs”, inset) after 1 6 h as a function of the concentration of fatty acid in the culture
medium. Square, OA; circles, DHA. Each point represents the mean ± 1 standard error
of the mean from n = 3 independent determinations of different cultures.
Summer 2015
24
Figure 5. The effect of DHA on cellular caspases. The activities of caspases (casp-) 3, 8,
and 9 in T27A cells after 16 h of culture in normal medium (control) and in medium
containing 0.61 mM DHA. The activity of caspase-9 is significantly (*, p < 0.05)
different from the control value. The activities of caspases -3 and -8 are not significantly
different from caspase-9 and are marginally significantly (0.05 < p < 0.1) different from
the controls. The data are presented as percent of activity found in control cells and
represent the means ± 1 standard error of the mean from n = 3 (caspase-3) or n = 4
independent assays using separate cell cultures.
inhibitor then Dl IA
Figure 6. Density of T27A cell cultures expressed as a percentage of that in control flasks
after 48 hours in the presence of medium containing no additions, 0.1% (vol/vol) DMSO
(carrier control), and medium enriched with 0.61 mM DHA. These are compared to the
densities of cell cultures exposed for 30 min to 10 (iM of an inhibitor specific to each
one of the indicated caspases before the 48 h exposure to medium containing 0.61 mM
DHA. Each bar represent the mean ± 1 standard error of the mean from n = 12 (control,
DMSO, and DHA) or n = 4 (inhibitors) independent cultures and assays. The bar labeled
with the asterisks is significantly (p < 0.05) different from the control value.
Washington Academy of Sciences
25
Discussion
T27A is a line of murine B lymphoblast cells with well-described
susceptibility to DHA-induced cell death (Zerouga et al. 1996; Kafrawy
et al. 1 998). Under normal growth conditions they possess very little DHA
(Table 1). During culture in media supplemented with DHA, T27A cells
incorporated considerable amounts of the fatty acid into phospholipids of
their plasma membrane (Table 1, Figure 1). Delivering exogenous
metabolites and drugs to cells as albumin conjugates is thought to simulate
physiological delivery conditions and buffers the availability of the
delivered substance. Other advantages of using albumin as a biological
carrier molecule are described elsewhere (Kratz 2008; Elsadek and Kratz
2012).
In phospholipids extracted from whole cells, DHA increased from
less than one-half of 1% of the total in control cells to over 10% of total
phospholipid fatty acids in cells cultured with supplemental DHA. In
purified plasma membrane preparations the percentage rose from close to
1% to over 18%. These results agree well with previous data from these
cells (Williams et al. 1998) and with reports showing that DHA has a
powerful effect on both the composition and structure of their plasma
membranes (Zerouga et al. 1996; Zerouga et al. 1997; Williams et al.
1998; Williams et al. 1999). These observations suggest that the
metabolism of fatty acids in these leukemic cells favors the non-random
incorporation of DHA into cell membranes with a preferential
incorporation of DHA into phospholipids of the plasma membrane.
Preferential incorporation of DHA into the plasma membranes of T27A
cells has been observed previously (Jenski et al. 1993; Pascale et al. 1993;
Williams et al. 1998; Williams et al. 1999).
Culture of T27A cells in DHA-enriched media caused a dose-
dependent decrease in cell density and cell viability, an increase in the
percent of cells exhibiting blebs, and the activation of cellular caspases.
OA did not induce these effects (Figures 2 and 3). We chose OA as the
control fatty acid for this study because it is the most abundant fatty acid
in many cell types, it is not toxic to T27A cells (Kafrawy et al. 1 998) and
because in other cells types and in model membranes it neither induces
apoptosis nor influences membrane raft function or structure (Kishida et
al. 2006; Shaikh et al. 2009; Shaikh et al. 2009a). Figures 2 and 3 show
Summer 2015
26
that when cells were cultured in media containing DHA at concentrations
above approximately 0.3 mM, their rate of proliferation was slowed and
significant numbers of cells died. The observation of induction of cell
death near 0.3 mM is consistent with what we and others have found in
this cell line (Zerouga et al. 1996; Williams et al. 1998; Williams et al.
1999), and may have implications for human health. In a study of healthy
men and women the serum concentration of DHA-phospholipids was
found to be near 0.15 mM. That concentration rose to over 0.35 mM after
six weeks of dietary supplementation with DHA capsules (Conquer and
Holub 1998). In a separate study of 234 healthy men, the mean serum
concentration of DHA-phospholipids was 0.18 mM and was elevated to
over 0.31 mM by similar DHA capsule supplementation (Grimsgaard et
al. 1997). These studies show that the DHA levels used to reduce the
growth and proliferation of mouse leukemia cells in vitro can be achieved
in humans by dietary manipulation.
At all concentrations of DHA examined, cells exhibited distinct
exvaginations or blebs on their plasma membranes (Figure 4). Though the
significance of these structures is not well understood, they are widely
recognized as a hallmark of apoptosis (Charras 2008). The definitive
indicator of apoptosis is the presence of active caspases (Galluzzi et al.
2011) and in these cells culture in medium containing 0.61 mM DHA
resulted in the activation of caspases -3, -8, and -9 (Figure 5). These
observations establish that DHA induces apoptosis in T27A cells.
In general, apoptosis can be triggered by two separate, but linked,
pathways: the intrinsic and extrinsic pathways (Portt et al. 2011; Galluzzi
et al. 2011). The intrinsic pathway originates with mitochondria and
involves the release from the intermembrane space of pro-apoptotic
molecules, particularly cytochrome c. The released cytochrome c initiates
a series of events that result in the conversion of inactive procaspase-9
into active caspase-9. Caspase-9 then activates caspase-3 which is
responsible for setting off the series of down-stream events characteristic
of apoptosis. The extrinsic pathway involves death receptors located in
the plasma membrane of the cell. Binding of an appropriate ligand to a
death receptor initiates an apoptotic cascade that begins with the
conversion of inactive procaspase-8 into active caspase-8. Depending on
the type of cell, caspase-8 then activates caspase-3 directly or indirectly
Washington Academy of Sciences
27
by converting the protein Bid into tBid which activates caspase-9 (Portt
et al. 2011; Galluzzi et al. 2011).
Caspases -3, -8, and -9 are all active in T27A cells after exposure
to DHA (Figure 5) and the inhibition of any one of them prevents the cells
from undergoing apoptosis (Figure 6). Since caspase-3 is an effector
caspase acting downstream of the initiator caspases -8 and -9, it appears
that a linear cascade of activation events occurs whereby one initiator
caspase activates the other (i.e. either caspase-8 activates caspase-9 or
vice versa) and the latter then activates caspase-3. In some cell types both
caspases -8 and -9 are able to activate caspases-3 directly (Slee et al. 1 999;
Peter and Krammer 2003), but apparently in T27A cells under the
conditions used here one of these caspases is unable to do so. It is possible
that one of the initiator caspases (-8 or -9) activates capsase-3, then
caspase-3 activates the remaining initiator caspase (Ozoren and El-Deiry
2003), but this is also not the case here because the activation of caspase-
3 initiates the irreversible stages of apoptosis (the execution pathway) and
thus the inhibition of the initiator caspase that was activated by caspases-
3 would not result in the rescue from cell death shown in Figure 5. The
data presented here suggest that one of the initiator caspases activates the
other yet is itself unable to activate caspase-3. These results are consistent
with T27A cells belonging to the type II group of apoptotic cells (Scaffidi
et al. 1998; Ozoren and El-Deiry 2002). In cells able to undergo type I
apoptosis, death receptor/ligand binding results in the direct activation of
effector caspases like caspase-3. Most cells undergo type II apoptosis, in
which death receptor/ligand binding is indirectly linked to the activation
of effector caspases through the mitochondrion-dependent pathways via
Bid and tBid (Scaffidi et al. 1998; Ozoren and El-Deiry 2002; Blanarova
et al. 2011). These observations are consistent with a pathway in T27A
cells in which DHA induces apoptosis by first triggering caspase-8 which
in turn activates caspase-9 to initiate the effector caspases.
Lipid rafts are dynamic and ephemeral laterally segregated
assemblies of the plasma membrane that are rich in sphingolipids,
cholesterol, and acylated and glycosylphosphatidylinositol (GPI)-
anchored proteins (Simons and Ikonen 1997; Lingwood and Simons
2010). Lipid rafts serve as important platforms for the regulation of cell
processes by confining and concentrating receptors and enzymes from the
Summer 2015
28
surrounding membrane. They represent another selective cellular
compartment that can co-localize and modulate the activities of these
proteins (Simons and Ikonen 1997; Lingwood and Simons 2010). Death
receptors, a subset of the tumor necrosis factor receptor superfamily, are
among those receptors that have been shown to be regulated by lipid rafts.
They include tumor necrosis factor receptor- 1 (TNF-R1, p55), death
receptor (DR) 3 (WSL-l/APO-3), DR4 (tumor necrosis factor-related
apoptosis-inducing ligand receptor- 1 [TRAIL-R1]), DR5 (TRAIL-
R2/APO-2), DR6 and CD95 (Fas/APO-1) (Ashkenazi and Dixit 1998;
Lavrik 2011). These receptors initiate extrinsic apoptosis after ligand
binding or ligand-independent clustering of receptors (Fumarola et al.
2001; Scheel-Toellner et al. 2004). Only when located within lipid rafts
do death receptors facilitate the activation of caspase-8 and down-stream
events leading to apoptosis. Death receptors do not activate caspase-8
when located in non-raft regions of the membrane (Xu et al. 2009; Gajate
et al. 2009; Blanarova et al. 2011).
Lipid raft dysfunction has previously been implicated in the DHA-
induced cell death of T27A cells (Williams et al. 1998; Williams et al.
1999). Other work has shown that DF1A alters the structure (Wassail and
Stillwell 2008), size (Chapkin et al. 2008; Rockett et al. 2012) and protein
composition (Rogers et al. 2010) of lipid rafts. Recently, Shaikh's group
has shown that DHA has profound effects on mammalian immune
function and that these effects arise from the influence of DHA on the
lipid rafts of B cells (Rockett et al. 2012; Gurzell et al. 2013). Other
evidence convincingly shows that DHA alters the raft-localization of
epidermal growth factor receptor (Schley et al. 2007; Rogers et al. 2010),
caveolin-1 (Li et al. 2007), toll-like receptors (Wong et al. 2009), the
major histocompatibility complex (MHC) class I proteins (Ruth et al.
2009; Shaikh et al. 2009), the signaling molecules SFK, Lck, Fyn, and c-
Yes (Stulnig et al. 1998; Stulnig et al. 2001; Chen et al. 2007), the
interleukin-2 receptor (Li et al. 2005), phospholipase D1 (Diaz et al.
2002), endothelial nitric oxide synthase (Li et al. 2007; Matesanz et al.
2010), and protein kinase C (Fan et al. 2004). Combined with the data
presented here, these observations suggest that DHA has an influence on
death receptor-mediated apoptosis via an action on lipid rafts. This
conclusion is reinforced by studies showing that a number of structurally
diverse anti-tumor agents selectively induce apoptosis in cancer cells by
Washington Academy of Sciences
29
triggering apoptosis thorough Fas-clustering in lipid rafts (Xu et al. 2009;
Mollinedo et al. 2010; Blanarova et al. 2011).
Conclusions
When T27A leukemic cells are cultured in media enriched with
DHA, the cells take up the fatty acid and incorporate it into their
membranes, particularly the plasma membrane. Culture in DHA-enriched
medium also causes cell death by inducing apoptosis. This induction of
apoptosis is caused by the initiation of the extrinsic apoptotic pathway and
with a linear activation of caspases in the sequence caspase-8, caspase-9,
then caspase-3. Coupled with previous observations by us and others, the
data suggest that the first step in DHA-induced cell death in T27A
leukemia cells involves the activation of plasma membrane-associated
death receptors by an influence of DHA on lipid rafts.
Acknowledgements
I am very grateful for the excellent assistance of Meagan E. O.
Smith and Marina Acocella. I also wish to thank Matt Anderson, Lisa
Fitzgerald-Miller, Kendra Model, Katrina Moncure, Melissa Moore,
Shaun Smith and Jeremy White. I thank Drs. J. Stribling and P. Erikson
for reading and commenting on drafts of the manuscript. The support of
the Department of Biological Sciences and of the Henson School of
Science and Technology of Salisbury University is acknowledged and
greatly appreciated.
Summer 2015
30
References
Ashkenazi A, Dixit VM. 1998. Death receptors: signaling and
modulation. Science 281:1305-1308.
Berger A, Roberts MA, Hoff B. 2006. How dietary arachidonic- and
docosahexaenoic- acid rich oils differentially affect the murine hepatic
transcriptome. Lipids Health Dis 5:10.
Blanarova OV, Jelinkova I, Szoor A, Skender B, Soucek K, Horvath
V, Vaculova A, Andera L, Sova P, Szollosi J, Hofmanova J, Vereb G,
Kozubik A. 201 1 . Cisplatin and a potent platinum (IV) complex-mediated
enhancement of TRAIL-induced cancer cells killing is associated with
modulation of upstream events in the extrinsic apoptotic pathway.
Carcinogenesis 32:42-51.
Blanckaert V, Ulmann L, Mimouni V, Antol J, Brancquart L, Chenais
B. 2010. Docosahexaenoic acid intake decreases proliferation, increases
apoptosis and decreases the invasive potential of the human breast
carcinoma cell line MDA-MB-231. Int J Oncol 36:737-742.
Bligh EG, Dyer WJ. 1 959. A rapid method of total lipid extraction and
purification. Can J Biochem Physiol 37:91 1-917.
Calder PC. 2012. Mechanisms of action of (n-3) fatty acids. J Nutr
142:592S-599S.
Chapkin RS, Kim W, Lupton JR, McMurray DN. 2009. Dietary
docosahexaenoic and eicosapentaenoic acid: emerging mediators of
inflammation. Prostag Leukotr Ess 81:187-191.
Chapkin RS, Wang N, Fan YY, Lupton JR, Prior IA. 2008.
Docosahexaenoic acid alters the size and distribution of cell surface
microdomains. Biochim Biophys Acta 1778:466-471.
Charras GT. 2008. A short history of blebbing. J Microsc 231:466-
478.
Chen W, Jump DB, Esselman WJ, Busik JV. 2007. Inhibition of
cytokine signaling in human retinal endothelial cells through modification
of caveolae/lipid rafts by docosahexaenoic acid. Invest Ophth Vis Sci
48:18-26.
Colas RA, Shinohara M, Dalli J, Chiang N, Serhan CN. 2014.
Identification and signature profiles for pro-resolving and inflammatory
Washington Academy of Sciences
31
lipid mediators in human tissue. Am J Physiol - Cell Physiol 307:C39-
C54.
Conover WJ, Johnson ME, Johnson MM. 1981 . A comparative study
of tests for homogeneity of variances, with applications to the outer
continental shelf bidding data. Technometrics 23:351-361.
Conquer JA, Holub BJ. 1998. Effect of supplementation with different
doses of DHA on the levels of circulating DHA as non-esterified fatty acid
in subjects of Asian Indian background. J Lipid Res 39:286-292.
Dalli J, Ramon S, Norris PC, Colas RA, Serhan CN. 2015. Novel
proresolving and tissue-regenerative resolvin and protectin sulfido-
conjugated pathways. FASEB J.
Diaz O, Berquand A, Dubois M, Di AS, Sette C, Bourgoin S, Lagarde
M, Nemoz G, Prigent AF. 2002. The mechanism of docosahexaenoic
acid-induced phospholipase D activation in human lymphocytes involves
exclusion of the enzyme from lipid rafts. J Biol Chem 277:39368-39378.
Eder K, Reichlmayr-Lais AM, Kirchgessner M. 1992. Studies on the
methanolysis of small amounts of purified phospholipids for gas
chromatographic analysis of fatty acid methyl esters. J Chromatog
607:55-67.
Elsadek B, Kratz F. 2012. Impact of albumin on drug delivery - New
applications on the horizon. J Control Release 157:4-28.
Fan YY, Ly LH, Barhoumi R, McMurray DN, Chapkin RS. 2004.
Dietary docosahexaenoic acid suppresses T cell protein kinase C theta
lipid raft recruitment and IL-2 production. J Immunol 1 73:6151-6160.
Fumarola C, Zerbini A, Guidotti GG. 2001. Glutamine deprivation-
mediated cell shrinkage induces ligand-independent CD95 receptor
signaling and apoptosis. Cell Death Differ 8:1004-1013.
Gajate C, Gonzalez-Camacho F, Mollinedo F. 2009. Involvement of
raft aggregates enriched in Fas/CD95 death-inducing signaling complex
in the antileukemic action of edelfosine in Jurkat cells. PLoS One 4:e5044.
Accessed on July 23 2015 and available at
http://joumals.plos. org/plosone/article?id=l 0. 1 37 1 /journal, pone.0005044.
Galluzzi L, Vitale I, Abrams JM, Alnemri ES, Baehrecke EH,
Blagosklonny MV, Dawson TM, Dawson VL, El-Deiry WS, Fulda S,
Gottlieb E, Green DR, Hengartner MO, Kepp O, Knight RA, Kumar S,
Lipton SA, Lu X, Madeo F, Malorni W, Mehlen P, Nunez G, Peter ME,
Summer 2015
32
Piacentini M, Rubinsztein DC, Shi Y, Simon HU, Vandenabeele P, White
E, Yuan J, Zhivotovsky B, Melino G, Kroemer G. 2011. Molecular
definitions of cell death subroutines: recommendations of the
Nomenclature Committee on Cell Death 2012. Cell Death Differ 19:107-
120.
Grimsgaard S, Bonaa KH, Hansen JB, Nordoy A. 1997. Highly
purified eicosapentaenoic acid and docosahexaenoic acid in humans have
similar triacylglycerol-lowering effects but divergent effects on serum
fatty acids. Am J Clin Nutr 66:649-659.
Gurzell EA, Teague H, Harris M, Clinthorne J, Shaikh SR, Fenton JI.
2013. DHA-enriched fish oil targets B cell lipid microdomains and
enhances ex vivo and in vivo B cell function. J Leukocyte Biol 93:463-
470.
Jenski LJ, Sturdevant LK, Ehringer WD, Stillwell W. 1993. Omega-3
fatty acid modification of membrane structure and function. I. Dietary
manipulation of tumor cell susceptibility to cell- and complement-
mediated lysis. Nutr Cancer 19:135-146.
Jenski LJ, Zerouga M, Stillwell W. 1995. Omega-3 fatty acid
containing liposomes in cancer therapy. Proc Soc Exp Biol Med 210:227-
233.
Kaduce TL, Awad AB, Fontenelle LJ, Spector AA. 1977. Effect of
fatty acid saturation on a-aminoisobutyric acid transport in Ehrlich ascites
cells. J Biol Chem 252:6624-6630.
Kafrawy O, Zerouga M, Stillwell W, Jenski LJ. 1998.
Docosahexaenoic acid in phosphatidylcholine mediates cytotoxicity more
effectively than other omega-3 and omega-6 fatty acids. Cancer Lett
132:23-29.
Kang KS, Wang P, Yamabe N, Fukui M, Jay T, Zhu BT. 2010.
Docosahexaenoic acid induces apoptosis in MCF-7 cells in vitro and in
vivo via reactive oxygen species formation and caspase 8 activation. PLoS
One 5 :e 10296. Accessed on July 23 2015 available at
http://journals.plos.org/plosone/article7icUl 0. 1 371 /journal. pone. 00 1 0296.
Kishida E, Tajiri M, Masuzawa Y. 2006. Docosahexaenoic acid
enrichment can reduce L929 cell necrosis induced by tumor necrosis
factor. BBA - Mol Cell Biol L 1761 :454-462.
Washington Academy of Sciences
33
Kratz F. 2008. Albumin as a drug carrier: Design of prodrugs, drug
conjugates and nanoparticles. J Control Release 1 32: 1 71 -1 83.
Lang I, Fick A, Schafer V, Giner T, Siegmund D, Wajant H. 2012.
Signaling active CD95 receptor molecules trigger co-translocation ot
inactive CD95 molecules into lipid rafts. J Biol Chem 287:24026-24042.
Lavrik I. 2011. Regulation of death receptor-induced apoptosis
induced via CD95/Fas and other death receptors. Mol Biol 45:150-155.
Li Q, Wang M, Tan L, Wang C, Ma J, Li N, Li Y, Xu G, Li J. 2005.
Docosahexaenoic acid changes lipid composition and interleukin-2
receptor signaling in membrane rafts. J Lipid Res 46:1904-1913.
Li Q, Zhang Q, Wang M, Liu F, Zhao S, Ma J, Luo N, Li N, Li Y, Xu
G, Li J. 2007. Docosahexaenoic acid affects endothelial nitric oxide
synthase in caveolae. Arch Biochem Biophys 466:250-259.
Lien EL. 2009. Toxicology and safety of DHA. Prostag Leukotr Ess
81:125-132.
Lingwood D, Simons K. 2010. Lipid rafts as a membrane-organizing
principle. Science 327:46-50.
Matesanz N, Park G, McAllister H, Leahey W, Devine A, McVeigh
GE, Gardiner TA, McDonald DM. 2010. Docosahexaenoic acid improves
the nitroso-redox balance and reduces VEGF-mediated angiogenic
signaling in microvascular endothelial cells. Invest Ophth Vis Sci
51:6815-6825.
Matta JA, Miyares RL, Ahern GP. 2007. TRPV1 is a novel target for
omega-3 polyunsaturated fatty acids. J Physiol 578:397-41 1.
Mischoulon D, Freeman MP. 2013. Omega-3 fatty acids in psychiatry.
Psychiat Clin N Am 36:15-23.
Mitchell DC, Niu SL, Litman BJ. 2003. DFlA-rich phospholipids
optimize G-Protein-coupled signaling. J Pediatr 143:S80-S86.
Mollinedo F, de la Iglesia-Vicente J, Gajate C, de Mendoza AEH,
Villa-Pulgarin JA, de Frias M, Roue G, Gil J, Colomer D, Campanero
MA, Blanco-Prieto MJ. 2010. In vitro and in vivo selective antitumor
activity of edelfosine against mantle cell lymphoma and chronic
lymphocytic leukemia involving lipid rafts. Clin Cancer Res 16:2046-
2054.
Summer 2015
34
Molnar J, Markovic G, Chao H, Molnar Z. 1969. Glycoproteins of
Ehrlich ascites carcinoma cells: Separation of plasma and endoplasmic
reticular membrane fragments. Arch Biochem Biophys 134:524-532.
Niu SL, Mitchell DC. 2005. Effect of packing density on rhodopsin
stability and function in polyunsaturated membranes. Biophys J 89: 1833-
1840. '
OzorenN, El-Deiry WS. 2002. Defining characteristics of Types I and
II apoptotic cells in response to TRAIL. Neoplasia 4:551-1557.
Ozoren N, El-Deiry WS. 2003. Cell surface death receptor signaling
in normal and cancer cells. Semin Cancer Biol 13:135-147.
Pascale AW, Ehringer WD, Stillwell W, Sturdevant LK, Jenski LJ.
1993. Omega-3 fatty acid modification of membrane structure and
function. II. Alteration by docosahexaenoic acid of tumor cell sensitivity
to immune cytolysis. Nutr Cancer 19:147-157.
Peter ME, Krammer PH. 2003. The CD95 (APO-l/Fas) DISC and
beyond. Cell Death Differ 10:26-35.
Pinot M, Vanni S, Pagnotta S, Lacas-Gervais S, Payet LA, Ferreira T,
Gautier R, Goud B, Antonny B, Barelli H. 2014. Polyunsaturated
phospholipids facilitate membrane deformation and fission by endocytic
proteins. Science 345:693-697.
Portt L, Norman G, Clapp C, Greenwood M, Greenwood MT. 2011.
Anti-apoptosis and cell survival: A review. BBA - Mol Cell Res 1813:238-
259.
Rockett BD, Teague H, Harris M, Melton M, Williams J, Wassail SR,
Shaikh SR. 2012. Fish oil increases raft size and membrane order of B
cells accompanied by differential effects on function. J Lipid Res 53:674-
685.
Rogers KR, Kikawa KD, Mouradian M, Hernandez K, McKinnon
KM, Ahwah SM, Pardini RS. 2010. Docosahexaenoic acid alters
epidermal growth factor receptor-related signaling by disrupting its lipid
raft association. Carcinogenesis 31:1523-1530.
Ruth MR, Proctor SD, Field CJ. 2009. Feeding long-chain n-3
polyunsaturated fatty acids to obese leptin receptor-deficient JCR:LA- cp
rats modifies immune function and lipid-raft fatty acid composition. Br J
Nutr 101:1341-1350.
Washington Academy of Sciences
35
Salem N, Kim H-Y, Yergey JA. 1986. Docosahexaenoic acid:
membrane function and metabolism. In: Simopoulos AP, Kifer RR,
Martin RE, editors. Health effects of polyunsaturated fatty acids in
seafoods. Orlando , Florida: Academic Press, p 263-317.
Scaffidi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli KJ,
Debatin KM, Krammer PH, Peter ME. 1998. Two CD95 (APO-l/Fas)
signaling pathways. EMBO J 17:1675-1687.
Scheel-Toellner D, Wang K, Assi LK, Webb PR, Craddock RM,
Salmon M, Lord JM. 2004. Clustering of death receptors in lipid rafts
initiates neutrophil spontaneous apoptosis. Biochem Soc Trans 32:679-
681.
Schley PD, Brindley DN, Field CJ. 2007. (n-3) PUFA alter raft lipid
composition and decrease epidermal growth factor receptor levels in lipid
rafts of human breast cancer cells. J Nutr 137:548-553.
Serhan CN, Dalli J, Colas RA, Winkler JW, Chiang N. 2014.
Protectins and maresins: New pro-resolving families of mediators in acute
inflammation and resolution bioactive metabolome. BBA - Mol Cell Biol
L 1851:397-413.
Serini S, Piccioni E, Merendino N, Calviello G. 2009. Dietary
polyunsaturated fatty acids as inducers of apoptosis: implications for
cancer. Apoptosis 14:135-152.
Shaikh SR. 2010. Diet-induced docosahexaenoic acid non-raft
domains and lymphocyte function. Prostag Leukotr Ess 82:159-164.
Shaikh SR, LoCascio DS, Soni SP, Wassail SR. Stillwell W. 2009.
Oleic- and docosahexaenoic acid-containing phosphatidylethanolamines
differentially phase separate from sphingomyelin. BBA - Biomembranes
1788:2421-2426.
Shaikh SR, Rockett BD, Salameh M, Carraway K. 2009a.
Docosahexaenoic acid modifies the clustering and size of lipid rafts and
the lateral organization and surface expression of MHC class I of EL4
cells. J Nutr 139:1632-1639.
Siddiqui RA, Harvey K, Stillwell W. 2008. Anticancer properties of
oxidation products of docosahexaenoic acid. Chem Phys Lipids 153:47-
56.
Summer 2015
36
Siddiqui RA, Shaikh SR, Sech LA, Yount HR, Stillwell W, Zaloga
GP. 2004. Omega 3-fatty acids: health benefits and cellular mechanisms
of action. Mini Rev Med Chem 4:859-871.
Simons K, Ikonen E. 1 997. Functional rafts in cell membranes. Nature
387:569-572.
Slee EA, Harte MT, Kluck RM, Wolf BB, Casiano CA, Newmeyer
DD, Wang HG, Reed JC, Nicholson DW, Alnemri ES, Green DR, Martin
SJ. 1999. Ordering the cytochrome c-initiated caspase cascade:
hierarchical activation of caspases-2, -3, -6, -7, -8, and -10 in a caspase-
9-dependent manner. J Cell Biol 144:281-292.
Sokal RR, Rohlf FJ. 1981. Biometry. New York: W.H. Freeman and
Company. 428 p.
Spector AA, Hoak JC. 1969. An improved method for the addition of
long-chain free fatty acid to protein solutions. Anal Biochem 32:297-302.
Stillwell W. 2008. Docosahexaenoic acid: a most unusual fatty acid.
Chem Phys Lipids 153:1-2.
Stillwell W, Shaikh SR, Zerouga M, Siddiqui RA, Wassail SR. 2005.
Docosahexaenoic acid affects cell signaling by altering lipid rafts. Reprod
Nutr Dev 45:559-579.
Stillwell W, Wassail SR. 2003. Docosahexaenoic acid: membrane
properties of a unique fatty acid. Chem Phys Lipids 126:1-27.
Stulnig TM, Berger M, Sigmund T, Raederstorff D, Stockinger H,
Waldhausl W. 1998. Polyunsaturated fatty acids inhibit T cell signal
transduction by modification of detergent-insoluble membrane domains.
J Cell Biol 143:637-644.
Stulnig TM, Huber J, Leitinger N, Imre EM, Angelisova P, Nowotny
P, Waldhausl W. 2001. Polyunsaturated eicosapentaenoic acid displaces
proteins from membrane rafts by altering raft lipid composition. J Biol
Chem 276:37335-37340.
Teague H, Ross R, Harris M, Mitchell DC, Shaikh SR. 2013. DHA-
fluorescent probe is sensitive to membrane order and reveals molecular
adaptation of DHA in ordered lipid microdomains. J Nutr Biochem
24:188-195.
Wassail SR, Stillwell W. 2008. Docosahexaenoic acid domains: the
ultimate non-raft membrane domain. Chem Phys Lipids 153:57-63.
Washington Academy of Sciences
37
Webb Y, Hermida-Matsumoto L, Resh MD. 2000. Inhibition of
protein palmitoylation, raft localization, and T cell signaling by 2-
bromopalmitate and polyunsaturated fatty acids. J Biol Chem 275:261-
270.
Williams EE, Jenski LJ, Stillwell W. 1998. Docosahexaenoic acid
(DHA) alters the structure and composition of membranous vesicles
exfoliated from the surface of a murine leukemia cell line. Biochim
Biophys Acta 1371:351-362.
Williams EE, May BD, Stillwell W, Jenski LJ. 1999.
Docosahexaenoic acid (DHA) alters the phospholipid molecular species
composition of membranous vesicles exfoliated from the surface of a
murine leukemia cell line. Biochim Biophys Acta 1418:185-196.
Williams EE, Somero GN. 1996. Seasonal-, tidal-cycle- and
microhabitat-related variation in membrane order of phospholipid
vesicles from gills of the intertidal mussel Mytilus californianus . J Exp
Biol 199:1587-1596.
Wong SW, Kwon MJ, Choi AMK, Kim HP, Nakahira K, Hwang DH.
2009. Fatty acids modulate Toll-like receptor 4 activation through
regulation of receptor dimerization and recruitment into lipid rafts in a
reactive oxygen species-dependent manner. J Biol Chem 284:27384-
27392.
Wren JJ. 1960. Chromatography of lipids on silicic acid . J Chromatog
4:173-195.
Xu ZX, Ding T, Haridas V, Connolly F, Gutterman JU. 2009. Avicin
D, a plant triterpenoid, induces cell apoptosis by recruitment of Fas and
downstream signaling molecules into lipid rafts. PLoS One 4:e8532.
Accessed on July 23 2015 available at
http://joumals.plos.org/plosone/article7idH0.1371/journal.pone.0008532.
Yamagami T, Porada CD, Pardini RS, Zanjani ED, Almeida-Porada
G. 2009. Docosahexaenoic acid induces dose dependent cell death in an
early undifferentiated subtype of acute myeloid leukemia cell line. Cancer
Biol Ther 8:331-337.
Zerouga M, Jenski LJ, Booster S, Stillwell W. 1997. Can
docosahexaenoic acid inhibit metastasis by decreasing deformability of
the tumor cell plasma membrane? Cancer Lett 1 19:163-168.
Summer 2015
38
Zerouga M, Stillwell W, Stone J, Powner A, Dumaual AC, Jenski LJ.
1996. Phospholipid class as a determinant in docosahexaenoic acid's
effect on tumor cell viability. Anticancer Res 16:2863-2868.
Bio
Eugene Williams is an educator and scientist who works in the area of cell
physiology. He is interested in how fishes acclimatize and adapt to
changes in temperature and in understanding the curious relationship
between the oils of cold-water fishes and cancers. Williams earned a Ph.D.
at Arizona State University and is currently a Professor of Biological
Sciences at Salisbury University.
Washington Academy of Sciences
39
A Nineteenth Century Historical Analysis of Game
Warden Efforts: Focus on Rabbits and Hares
Kelsey Gilcrease
South Dakota School of Mines and Technology
Abstract
Leporids (rabbits and hares) were widely sought-after game animals to
many people in the nineteenth century. But how often were offenses to
the game laws caught? That answer depends on the number of wardens,
the amount of prosecuted leporid offenses (as compared to other
offenses), and how complex it was to catch an offense. The aim of this
paper is to determine the types of offenses that game wardens enforced,
the number of prosecuted leporid offenses, the specific types of leporid
offenses in two states — New Jersey and Massachusetts — and the New
Jersey counties where leporid enforcements occurred. This investigation
uncovered three key findings: (1) The more uniformity of time spent
between wardens on offenses could be a success factor in catching
leporid offenses; (2) There was a correlation between the number of
wardens who cited leporid offenses and the number of counties involved
with the leporid offenses; and (3) Outside of those years, there was a
disproportionate number of leporid offenses when correlating for the
numbers of wardens and numbers of counties. Furthermore, the results
of this investigation offer implications toward our understanding of past
leporid conservation, most notably findings related to the uniformity of
the number of wardens prosecuting leporid offenses and the years when
prosecuted leporid offenses were prominent in the nineteenth century.
Introduction
The historical emphasis of managing wildlife was largely synonymous
with managing game species and predators (Bolen and Robinson, 2003, pp.
20, 183). For example, the first game laws in North America occurred in
1639 to close the white-tailed deer hunting season for six months (Bolen
and Robinson, 2003). Protection of game is critical in efforts to conserve
wildlife because of the need to understand biology at the organismal level
and to conserve habitats (Willis et al., 2008).
However, the very beginnings of enforcement of wildlife laws in the
United States tell a rugged story. Beginning with colonial times, game law
enforcers intended that private citizens act as “advisors'” regarding game
laws to ensure that fellow hunters were following the laws (Lund, 1980),
Summer 2015
40
which would mostly help to inform the privileged class in how to protect or
manage game resources on their own lands. Yet if private citizens were
enforcing the game laws, how effective were they at protecting wild game?
In historical analyses of wildlife enforcement during the nineteenth century,
Lund (1980), Tober (1981), and Stockdale (1993) indicated that
enforcement of laws pertaining to wildlife was weak or non-existent, since
it was the people who were conducting the enforcements and this perception
was criticized by the local townspeople. Early wildlife laws too, ignored the
bag limit, further making enforcement of the laws difficult (Lund, 1980).
As enforcement of laws and compliancy is important to the
conservation of wildlife, the historical nineteenth-century rationale or
prioritization of the protection of leporids (rabbits and hares) in New Jersey
and Massachusetts (the two states that are the focus of this study) appears
unclear1. What offenses did the wardens focus their time on? And which
counties were enforcing the leporid laws? These questions lead to further
questions as to whether certain counties in New Jersey and Massachusetts
reported more leporid offenses than others, and what factors were most
strongly associated with whether and how game law violations were
reported.
Furthermore, research regarding wildlife conservation officers is
limited (Archbold, 2012) so there is a need for more of a historical
underpinning on the number of conservation offenses, the number of
wardens, and the county distribution of the offenses.
These questions are important because leporids were often hunted
for food ( Omaha Daily Bee , 1887, Si. Paul Daily Globe , 1887) at a time
when some North American leporid populations started declining. Let’s
look at the two states which, again, are the focus of this study - New Jersey
and Massachusetts.
The leporids of New Jersey include the Eastern cottontail
(Sylvilagus floridanus ), introduced species of the European hare ( Lepus
europaeus), the black-tailed jackrabbit (Lepus calif or nicus), the white-
tailed jackrabbit (Lepus townsendii ) (State of New Jersey, 2004) and, at one
time, the snowshoe hare (Lepus americanus ) (Rhoads, 1903). However,
current records indicate no presence of the snowshoe hare in New Jersey
(Murray and Smith, 2008).
Washington Academy of Sciences
41
The leporids of Massachusetts also include the Eastern cottontail,
snowshoe hare, and black-tailed jackrabbit (Massachusetts Executive
Office of Energy and Environmental Affairs, 2014). In Massachusetts, the
New England cottontail ( Sylvilcigus transUionalis ) currently has a
conservation status of candidate species (USFWS, 2015).
The purpose of this paper is to determine the types of leporid
offenses; the number of prosecuted leporid offenses; the types of offenses
game wardens enforced; and the counties that prosecuted the enforcements
— rather than to decipher why game wardens focused on specific leporid
species offenses.
Methods
Based on the Annual Reports of the Board of Fish and Game
Commissioners in New Jersey (1894-1899) and the Reports of the
Commissioners on Inland Fisheries and Game in Massachusetts (1889-
1 899), I categorized the nineteenth-century wildlife offenses for those two
states into these eight groups: fish, illegal fishing, lobster, Sunday offenses,
pollution, illegal game, trespassing, and any offenses related to game
generally, such as squirrels, deer, ducks, birds, and leporids. Table 1
presents a description of each of these eight categories.
Each year from 1894-1899 for New Jersey and 1889-1899 for
Massachusetts, I recorded the:
• total number of wardens who were holding a warden status in New
Jersey and Massachusetts,
• total number of offenses2,
• number of fish, illegal fishing, Sunday hunting, pollution, illegal
game, trespassing, and other game offenses (including the number
of wardens who cited leporid offenses),
• number of counties with leporid offenses,
• names of the wardens who cited leporid offenses, and
• counties in which the offenses occurred.
The Simpson’s E or Even-ness index (widely employed for
biodiversity studies), was used to determine the “even-ness” of the
prosecuted leporid offenses and how “evenly” the leporid offenses occurred
in each county (National Center for Ecological Analysis and Synthesis
Summer 2015
42
(NCEAS), 2014). Simpson’s E can range from 0 to 1, with 1 being the
highest uniformity (NCEAS, 2014). Because there was usually only one
warden spending time on leporid offenses in Massachusetts, the even-ness
index had to include the total number of wardens available, regardless of
whether all the wardens worked on game such as leporids.
Table 1. Offenses pursued in New Jersey and Massachusetts and
characterization of the offenses
Regarding the county data: In New Jersey only the years 1 894-1895,
1898, and 1899 contained county data. Massachusetts county data were not
applicable because county data were sparse, and due to the nature of the
historical documents, it was difficult to decipher where the offenses
originated. For the New Jersey data, correlation analysis was used to
determine the relationship between the number of wardens who cited
leporid offenses and the number of counties involved with leporid offenses
in the state of New Jersey.
Washington Academy of Sciences
43
Results
Findings on New Jersey
In New Jersey, fishing and illegal fishing activities were the more
commonly-reported offenses in 1894-1895 (49% of the overall time). In
1896, Sunday hunting and offenses with song birds were the more
commonly reported offenses (63% of the overall time). In 1897, illegal
fishing activities and offenses with song birds were more commonly
reported offenses in the state (60% of the time). By 1898 and 1899, Sunday
hunting and offenses with song birds were more commonly reported (57%
of the time for both years).
The above overall prosecuted offenses provide the metrics by which
annual percentages of prosecuted leporid offenses are calculated for the
state of New Jersey during those same years (Figure 1).
Percentage of leporid offenses in
New Jersey
Figure 1. New Jersey: Percentage of leporid offenses, among other types of offenses.
In New Jersey, the number of wardens each year did not
uniformly reflect the number of leporid offenses. The highest number of
wardens involved with leporid offenses occurred in 1896 with 14 wardens.
At the same time, 1896 had the lowest percentage of wardens reporting
leporid offenses (Figure 1). Simpson E values related to the even-ness of
wardens prosecuting leporid offenses in each year are reported in Table 2.
Summer 2015
44
Table 2. New Jersey: Even-ness of wardens prosecuting leporid offenses.
In New Jersey, the number of counties in which leporid offenses
were prosecuted each year did not uniformly reflect the total leporid
offenses in the state. The most leporid offenses (relative to other offenses)
occurred in 1 898 and 1 899 (Figure 1); however, the least number of counties
that cited leporid offenses occurred in 1899 (Figure 2). The year 1898 had
the most even number of wardens spending time on leporid offenses (Table
2) and 1898 is also significant for the even-ness of the counties that cited
leporid offenses (see Table 3, which shows Simpson E values for the even-
ness of counties with prosecuted leporid offenses in 1894-1895, 1898, and
1 899). The year 1 899 was the least uniform in terms of counties with leporid
offenses (Table 3), with Bergen County carrying over half of the total
leporid offenses (Figure 2).
The number of wardens who prosecuted leporid offenses
correlates strongly to the number of counties in which offenses were
prosecuted in New Jersey. Correlation co-efficient analysis showed r =
0.989 association between the number of wardens who cited leporid
offenses and the number of counties involved with the leporid offenses. In
1 896 and 1 898, the number of wardens who worked on leporid offenses was
more than 50% of all wardens (Figure 3); however, the most counties
prosecuting leporid offenses occurred in 1898 with twelve counties
involved (Figure 4). In the years 1894-1895, eight counties were involved
in leporid offenses (Figure 5).
Washington Academy of Sciences
45
New Jersey counties with leporid offenses in
1899
cu
County
Figure 2. New Jersey counties prosecuting leporid offenses, 1899.
Table 3. New Jersey: Even-ness of counties prosecuting leporid offenses.
Wardens prosecuting leporid
offenses in
New Jersey
0.7 i
1894 1895 1896 1897 1898 1899
Figure 3. New Jersey: Number of wardens who prosecuted leporid
offenses, expressed as a percentage of the total number of wardens.
Summer 201 5
46
The most prominent year for prosecuting leporid offenses in New
Jersey was 1898, based upon the previously discussed percentage of
wardens and percentage of prosecuted leporid offenses in New Jersey
(Figures 3 and 1), and also based upon the number of wardens prosecuting
leporid offenses being the most uniform (Table 2).
The most un-even number of wardens prosecuting leporid offenses
occurred in 1897, given that, for example, Warden Dunham carried 44% of
the total leporid offenses. Also, in 1897, there was a decrease (from the
previous year) in the number of wardens contributing to leporid offense
prosecutions (Figure 3).
New Jersey counties with leporid offenses in
1898
County
Figure 4. New Jersey counties prosecuting leporid offenses, 1898.
New Jersey had various types of prosecuted leporid offenses, with
the majority of the offenses involving killed rabbits and the lowest number
of offenses for selling rabbits (Figure 6). The methods used by hunters in
the killed-rabbit offenses was not mentioned.
Washington Academy of Sciences
47
New Jersey counties with leporid
offenses in 1894-1895
I I I I I
i r
County
.1
Figure 5. New Jersey counties prosecuting leporid offenses, 1894-1895.
Type of leporid offenses in New
Jersey
14
12
10
8
6
4
2
0
xO^ <A° A
^ of ,<$ 'A
&
* <?
4
■ 1894
1895
■ 1896
■ 1897
K 1898
1899
1900
Figure 6. New Jersey: Types of leporid offenses (killed, possession of,
snared, snooded, use of ferret, trapped, or sale of leporids).
Summer 2015
48
Findings on Massachusetts
In Massachusetts, the number of wardens reflects more prosecuted
leporid offenses. The most number of wardens dealing with leporid offenses
in Massachusetts was two; this occurred in 1894. The year 1894,
consequently, had the most prosecuted leporid offenses of any year from
1889 to 1899 (Figures 7 and 8).
Percentage of wardens
prosecuting leporid offenses in
Massachusetts
Figure 7. Massachusetts: Number of wardens who prosecuted for leporid
offenses, expressed as a percentage of the total number of wardens.
Figure 8. Massachusetts: Percentage of prosecuted leporid offenses,
relative to other types of offenses.
Washington Academy of Sciences
49
The year 1897 had the least uniform percentage of leporid offenses
in Massachusetts. As indicated in 'fable 4, for Massachusetts, 1 897 had the
lowest percentage of wardens dealing with leporid offenses along with the
low percentage of prosecuted leporid offenses previously shown in Figures
7 and 8.
Table 4. Massachusetts: Even-ness of prosecuted leporid offenses by year.
When wardens had more uniform leporid offense prosecutions, they
correlated to the number of wardens and percentage of prosecuted offenses
for leporid offenses. Massachusetts was not very uniform with regard to the
percentage of prosecuted leporid offenses being far from 1 with regard to
Simpson’s E (Table 4). During the years 1895-1899 (with the exception of
1898 with no leporid offenses), there was only one warden who prosecuted
leporid offenses in Massachusetts. Even though only one warden prosecuted
leporid offenses during those years, the year 1894 had the most uniform
prosecuted leporid offenses (Table 4) and this reflected the highest
percentage of wardens dealing with leporid offenses and the highest
percentage of leporid offenses.
As previously noted, the year 1897 was the least uniform with
prosecuted leporid offenses in Massachusetts (Table 4). The year 1897 also
reflected the lowest percentage of the number of wardens dealing with
leporid offenses in Massachusetts and the lowest percentage of prosecuted
leporid offenses.
Analysis and Implications
For both New Jersey and Massachusetts, 1 897 was the least uniform
in terms of prosecuted leporid offenses (Tables 2 and 4). For New Jersey,
this did not mean that the lowest uniformity in 1897 reflected the lowest
percentage of wardens or prosecuted leporid offenses. For Massachusetts it
did — and when wardens had more uniform annual prosecutions for leporid
offenses, there was also the highest percentage of wardens and highest
percentage of prosecuted leporid offenses.
Summer 2015
50
On the one hand, the year 1894 was the most prominent year for
catching leporid offenses in Massachusetts; however, Massachusetts'
“highest” point was New Jersey’s “lowest” point (Figures 3 and 7). On the
other hand, Massachusetts had a higher percentage of prosecuted leporid
offenses than New Jersey did in 1894, and the following years were
equivalent (Figures 1 and 8). This may be because the total numbers of
offenses were higher in New Jersey than in Massachusetts. Massachusetts
focused more on ferret offenses (a total of 21 offenses) and three offenses
with a snare (Figure 9).
Figure 9. Massachusetts: Types of leporid offenses (killed, possession of, snared,
snooded, use of ferret, trapped, or sale of leporids).
Conclusions and Additional Implications
The results from this research demonstrated a strong correlation
(r = 0.98) between the number of wardens that cite leporid offenses and the
number of counties involved with the leporid offenses in New Jersey.
However, there could be several variables that impinge on the relationship
between the number of wardens who cited leporid offenses and the number
of counties involved with the leporid offenses. For example, it is unknown
how often the wardens traveled between counties or if the warden remained
in the county in which he lived. In fact, Warden Post of Somerset County
wrote that he conducted a majority of his inspections near where he lived in
Somerset County (Annual Report of the Board of Fish and Game
Washington Academy of Sciences
51
Commissioners of the State of New Jersey, 1894). Likewise, some of the
counties had more than one warden enforcing laws and in some cases, there
were two wardens. Furthermore, it was difficult to discern if the wardens
actively looked for leporid offenses or if they happened to come across
offenses.
It would be useful to understand why, for New Jersey in 1 899, there
was a higher percentage of prosecuted leporid offenses yet the least number
of counties involved. It was also the most disproportionate year in terms of
the number of leporid offenses occurring in the counties, with Bergen
County leading the way with over 50% of the leporid offenses in 1 899. This
result may be because it was the second lowest uniformity for prosecuted
leporid offenses.3
Further research could try to determine why some years had more
offenses related to fishing, and then considerably fewer fishing offenses in
1 898 and 1 899. This investigation revealed that hunting on Sunday and song
bird offenses were relatively common prosecutions for every year studied.
At this point, it seems related to comment on why more prosecutions
may have focused on Sunday hunting and song bird offenses. Lund (1980)
and Lueck (1995) provided suggestions in terms of facilitation and
economics as to why some offenses were enforced more than others. Lund
(1980) described various methods early legislators used to make
enforcement easier (see Lund, 1980, for more details). For example, Lund
(1980) described that closed seasons was easy to enforce ( i.e ., a closed
season could be equivalent to no hunting on Sunday). Lueck (1995)
discussed more of an economical approach and suggested that by not having
hunting on Sunday, land could be used for other productive uses and the
prohibitive hunting on Sunday was a chance to reduce contracting costs.
Also, Lund ( 1 980) proposed that it was easier to make a prosecution
when game was being sold, rather than witnessing the offense in the
countryside. It is surprising that prosecutions for selling leporid meat were
not more prevalent (i.e., this study found that it was the killed leporid or use
of a ferret that were more dominant offenses). It would be insightful to note
the kinds of factors that would make leporid offenses, specifically, easier to
discover.
Furthermore, the years 1894-1899 seemed to have consistently
fewer pollution, illegal game, and trespassing offenses. There may have
Summer 2015
52
been a lower number of offenses related to illegal game because it was a
somewhat vague offense — citing only “game” and not a specific species.4
However, the lower number of prosecuted trespassing offenses may be
because trespassing was an offense that could be handled more self-
sufficiently with a landowner (i.e., not needing a warden) since there was
the presumption that — unless owners posted notices saying hunting was
not allowed on the land — people could continue to hunt (Lund, 1976).
One reason Massachusetts may not have had any leporid offenses in
1 898 may be because the report expressed an abundance of leporids (Report
of the Commissioners on Inland Fisheries and Game, 1898), so perhaps the
wardens felt that leporid enforcement did not need to be as heavily enforced
that year. For example. Wardens Smith and Manly reported observing the
illegal use of ferrets during hunting, but also reported that they did not catch
anyone in the act of crime (Report of the Commissioners on Inland Fisheries
and Game, 1898).
Finally, as noted, this investigation revealed that the number of
wardens did not always reflect more prosecuted leporid offenses in New
Jersey although it did for Massachusetts. Further research could focus on
why more wardens did not lead to more leporid prosecutions in New Jersey.
It should also be pointed out that future research efforts will need to
take into consideration any new findings regarding leporid population
cycles. That is, some leporid populations are believed to cycle in 10-year
intervals which could affect the types of variables being measured here. For
example, the snowshoe hare species appears to exhibit 10-year cycles,
although other leporids have not been researched as extensively in this
regard. The possible display of a 1 0-year population flux by other leporid
species, however, is a consideration for future related research.
Washington Academy of Sciences
53
Notes
1 New Jersey and Massachusetts were the states highlighted in this study because the data
were accessible.
2 In the reports, there was no distinction between offenses and prosecutions. Even when a
case was dismissed, it was still classified as an offense.
3 The study did not control for Bergen County statistically and then recalculate the
Simpson’s E because the goal was to examine the outstanding counties, overall, not to
do a county by county comparison.
4 In terms of whether the referenced reports included or did not include leporids,
unfortunately, in many cases, historical articles are the only resources available and,
in many instances, the label "‘game” may have meant deer, ducks, etc., which were
inexplicably lumped into one category.
References
Archbold, C. A. 2012. Policing'. A text/reader. SAGE Publications. Thousand Oaks, CA.
Bolen, E. G. and W. L. Robinson. 2003. Wildlife Ecology and Management 5th Edition.
Prentice Hall. Upper Saddle River, NJ.
Commonwealth of Massachusetts. 1889. Report of the Commissioners on Inland
Fisheries and Game. 1889. Boston, MA.
Commonwealth of Massachusetts. 1890. Report of the Commissioners on Inland
Fisheries and Game. 1890. Boston, MA.
Commonwealth of Massachusetts. 1 892. Report of the Commissioners on Inland
Fisheries and Game. 1892. Boston, MA.
Commonwealth of Massachusetts. 1 893. Report of the Commissioners on Inland
Fisheries and Game. 1893. Boston, MA.
Commonwealth of Massachusetts. 1894. Report of the Commissioners on Inland
Fisheries and Game. 1894. Boston, MA.
Commonwealth of Massachusetts. 1895. Report of the Commissioners on Inland
Fisheries and Game. 1 895. Boston, MA.
Commonwealth of Massachusetts. 1896. Report of the Commissioners on Inland
Fisheries and Game. 1 896. Boston, MA.
Commonwealth of Massachusetts. 1897. Report of the Commissioners on Inland
Fisheries and Game. 1897. Boston, MA.
Commonwealth of Massachusetts. 1898. Report of the Commissioners on Inland
Fisheries and Game. 1898. Boston, MA.
Commonwealth of Massachusetts. 1899. Report of the Commissioners on Inland
Fisheries and Game. 1899. Boston, MA.
Summer 2015
54
Commonwealth of Massachusetts. Executive Office of Energy and Environmental
Affairs. 2014. Hunting of Cottontail rabbit, Snowshoe hare, and Jackrabbit.
Available at: <http://www.mass.gov/eea/agencies/dfg/dfw/laws-regulations/plain-
lang-sum/hunting-of-cottontail-rabbit-snowshoe-hare-and-jackra.html >. Retrieved
on August 3, 2014.
Fowler, J., L. Cohen, and P. Jarvis. 1998. Practical Statistics for Field Biology’. John
Wiley and Sons. West Sussex, England.
Lueck, D. 1995. Property Rights and the Economic Logic of Wildlife Institutions.
Natural Resources Journal 35: 625-670.
Lund, T. A. 1976. Early American Wildlife Law. New York University Law Review 5 1 :
703-730.
Lund, T. A. 1980. American Wildlife Law’. University of California Press. Berkeley, CA.
Murray, D. and A. T. Smith. 2008. Lepus americanus. The IUCN Red List of Threatened
Species. Version 2014.2. <www.iucnredlist.org>. Retrieved on 06 August 2014.
National Center for Ecological Analysis and Synthesis (NCEAS). 2014. Available at:
<https://groups.nceas.ucsb.edu/sun/meetings/calculating-evenness-of-habitat-
distributions>. Retrieved August 8, 2014.
Omaha Daily Bee. 1887. March 7, 1887 page 4 Available at:
<http://chroniclingamerica.loc.gov/lccn/sn9902 1 999/1 887-03-07/ed- 1/seq-
4/#date 1 = 1 836&index=0&rows=20&words=destitute+rabbits&searchType=basic
&sequence=0&state=&date2=1900&proxtext=rabbits+destitute&y=0&x=0&dateF
ilterType=yearRange&page=l>.
Rhoads, S. N. 1903. The Mammals of Pennsylvania and New Jersey: A Biographic,
Historic and Descriptive Account of the Furred Animals of Land and Sea, both
living and extinct, known to have existed in these states. Wickersham Printing
Company. Lancaster, PA.
State of New Jersey. 1 894. Annual Report of the Board of Fish and Game Commissioners
of the State of New Jersey. 1 894. Trenton, NJ.
State of New Jersey. 1895. Annual Report of the Board of Fish and Game Commissioners
of the State of New Jersey. 1 895. Trenton, NJ.
State of New Jersey. 1 896. Annual Report of the Board of Fish and Game Commissioners
of the State of New Jersey. 1 896. Trenton, NJ.
State of New Jersey. 1 897. Annual Report of the Board of Fish and Game Commissioners
of the State of New Jersey. 1897. Trenton, NJ.
State of New Jersey. 1 898. Annual Report of the Board of Fish and Game Commissioners
of the State of New Jersey. 1898. Trenton, NJ.
State of New Jersey. 1 899. Annual Report of the Board of Fish and Game Commissioners
of the State of New Jersey. 1 899. Trenton, NJ.
State of New Jersey. 2004. Available at
<http://www.state.nj.us/dep/fgw/chkmamls.htm>. Retrieved on August 3, 2014.
Washington Academy of Sciences
55
St. Paul Daily Globe. 1887. February 18, 1887. Available at
<http://chroniclingamerica.loc.gov/lccn/sn90059522/1887-02-l 8/ed-l/seq-
2/#date 1 = 1 836&index=4&date2= 1 922&searchType=advanced&language=&sequ
ence=0&words=destitute+rabbits&proxdistance=50&state=&rows=20&ortext=&p
roxtext=rabbits+destitute&phrasetext=&andtext=&dateFilterType=yearRange&pa
ge=l>. Retrieved on August 9, 2014.
Stockdale, M. 1993. English and American Wildlife Law: Lessons from the Past.
Proceedings of the Annual Conference of the Southeastern Association of Fish and
Wildlife Agencies 47: 732-739.
Tober, J. A. 1981. Who Ch\’ns the Wildlife? The Political Economy of Conservation in
Nineteenth-Century America. Greenwood Press. Westport, CT.
United States Fish and Wildlife Service. 2015. New England Cottontail (Sylvilagus
transitionalis). Available at
<http://ecos.fws.gov/speciesProfile/profile/speciesProfile.action?spcode=A09B>.
Retrieved on August 10, 2015
Willis D., C. Scalet, and L. D. Flake. 2008. Introduction to Wildlife and Fisheries. W. H.
Freeman and Company. New York, NY.
Bio
Kelsey Gilcrease is a biology lab and ecology instructor at the
South Dakota School of Mines and Technology, Department of Chemistry
and Applied Biological Sciences. Her main research is focused on leporid
conservation in North America, biogeography, and conservation biology
histories in the nineteenth and early twentieth centuries. She may be
contacted at Kelsev.Gilcrease@sdsmt.edu.
Summer 2015
56
Washington Academy of Sciences
57
Uranus and Neptune Revisited
Sethanne Howard
USNO, retired
Abstract
Uranus and Neptune are two planets not known in ancient times. Once
discovered, however, astronomers were eager to obtain their vital
statistics. Of course once Voyager flew by them everyone had the
information, but before there was Voyager, there were many attempts
to measure things like the rotational period, the mass, the brightness,
etc. This is the story of the two people who obtained the rotational
period of both planets before Voyager got there. They confirmed that
the spin of Uranus is retrograde and that of Neptune direct. Uranus
rotates on its side. Their estimates for the periods of rotation are, for
Uranus, 24 ±3 hr., and for Neptune, 15 ±3 hr.
Introduction
For most of human history humanity knew of only five planets (plus
our own): Mercury, Venus, Mars, Jupiter, and Saturn. These are the
planets that are visible to the naked eye at various times during the year.
We learned relatively recently that there are two other planets in our Solar
System: Uranus and Neptune. Note the proper pronunciation of Uranus
(accent on the first syllable).
First a little history on Uranus and Neptune. The question being
what did we know before the Voyager flyby of Uranus.
Uranus had been observed on many occasions before its
recognition as a planet, but it was generally mistaken for a star. Then Sir
William Herschel observed the planet qua planet on March 13, 1781. This
was the first planet added to the Solar System since the dawn of history.
See Figure 1 for a drawing of the telescope he used. He decided to name
the new planet Georgium Sidus (George’s Star), in honor of his new
patron, King George III of England. As one might expect, this was not
popular outside England. Bode (a German astronomer) opted for Uranus ,
the Latinized version of the Greek god of the sky, Ouranos (the only
planet with a name of Greek origin). Bode argued that just as Saturn was
the father of Jupiter, the new planet should be named after the father of
Saturn. Ultimately, Bode’s suggestion became the most widely used, and
became universal in 1850 when Her Majesty’s Nautical Almanac Office,
the final holdout, switched from using Georgium Sidus to Uranus.
Summer 2015
58
Figure 1 — Discovery telescope for Uranus
Uranus is the seventh planet from the Sun. It has the third-largest
radius and fourth-largest mass in the Solar System. It revolves around the
Sun once every 84 Earth years. Uranus has a ring system and numerous
moons. The Uranian system has a unique configuration among the planets
because its axis of rotation is tilted sideways, nearly into the plane of its
revolution about the Sun. Its north and south poles therefore lie where
most other planets have their equators.
Uranus’s orbital elements (the shape of its orbit) were first
calculated in 1783 by Pierre-Simon Laplace.' Over time, discrepancies
began to appear between the predicted and observed orbits, and in 1841,
John C. Adams" first proposed that the differences might be due to the
gravitational tug of an unseen planet."1 In 1 845, Urbain Le Verrierlv began
his own independent research into Uranus’s orbit. On September 23,
1 846, Johann G. Gallev was the first to see the new planet close to the
position predicted by Le Verrier.vl
There was considerable controversy over the name for the new
planet. At first, Neptune was simply called “the planet exterior to Uranus”
or “Le Verrier’s planet.” However, eventually the name Neptune , Roman
god of the sea, was accepted.
Neptune is the eighth and farthest planet from the Sun in the Solar
System. It is the fourth-largest planet by diameter and the third-largest by
Washington Academy of Sciences
59
mass. It revolves around the Sun once every 164.8 Earth years. Like
Uranus, Neptune has a ring system and several moons.
Uranus - the Details
Uranus was known to be a bit strange. It was already suspected
that its axis of rotation was off kilter. Most of the Solar System planets
have rotation axes that are close to perpendicular to the plane of the orbit.
The Earth, for example, has an axis tilted only 23.5° off perpendicular.
Uranus, on the other hand, was thought to have a rotation axis almost in
its orbital plane. No one was quite sure of this, though.
Its major moons are Ariel, Umbriel, Titania, Oberon, and Miranda
- names taken from Shakespeare. There are about 27 moons known.
The internal structure of Uranus is shown in Figure 2. The wildly
off-center magnetic field is shown in Figure 3.
Figure 2 — Structure of Uranus
The planet is thought to have a very small central almost rocky
core, surrounded by a plasma ocean, surrounded in turn by an atmosphere
with lots of hydrogen, helium, and methane (CEE). It is the methane that
Summer 2015
60
makes Uranus appear cyan in color. Note in Figure 3 that the Magnetic
North Pole points “downward” - below the orbital plane.
The magnetic field is not centered on or near the center of the
planet. This is quite unusual. This unusual geometry results in a highly
asymmetric magnetosphere. By comparison, the magnetic field of Earth
is roughly the same at either pole, and its “magnetic equator” is roughly
parallel with its geographical equator.
Figure 3 — magnetic field for Uranus (left) and Neptune (right)
Neptune - the Details
Figure 4 shows the structure of Neptune. Although smaller than
Uranus as seen from the Earth, when seen with a large telescope it is
visible as a disk.
Voyager found the axial tilt of Neptune to be 28.32° - similar to
the Earth's tilt. Triton is its major moon - very large as moons go. Unlike
other large planetary moons in the Solar System, Triton has a retrograde
orbit, indicating that it was captured rather than formed in place. There
are about 14 known moons for Neptune. The magnetic field of Neptune
(Figure 3) is also a bit off center although not as much as Uranus.
Never visible to the naked eye, Neptune requires a 4 meter class
telescope to capture its spectra, and a 50 inch telescope to work in the near
infrared part of the spectrum. To me it is a beautiful planet because its
color is a deep, rich blue. The atmosphere is mainly hydrogen and helium
with trace amounts of CFU that contribute to that beautiful color.
Washington Academy of Sciences
61
Planetary Rotation
All planets rotate. The Earth rotates about its axis once a day. This
is how we define a "day’. Planets revolve around the Sun and rotate about
their individual axes. It is fairly straightforward to obtain the rotation
periods (i.e., the length of the planetary day) of the five naked eye planets
- one simply watches them. We can’t watch the more distant Uranus and
Neptune. It takes a large telescope to determine their rotational periods.
Figure 4 — The internal structure of Neptune:
1 . Upper atmosphere, top clouds
2. Atmosphere consisting of hydrogen, helium, and CH4 gas
3. Mantle consisting of water, ammonia, and CH4 ices
4. Core consisting of rock (silicates and nickel-iron)
By using various techniques people tried to determine the rotation
periods for Uranus and Neptune. A visual technique means watching the
planet as it spins. This is rather like watching the Great Red Spot on
Jupiter as Jupiter rotates: once around is a ‘day’ on Jupiter. Theory means
that the rotation period is derived from planetary theory (using the mass
and shape of the planet to derive its period), not by using a telescope.
Photometry means that a telescope is used with a filter in selected
wavelength bands ( e.g ., a color like infrared) to measure changes in the
light from the planet. A regular and repeatable change in the light can
Summer 2015
62
represent the length of the planetary day. The spectra technique means
that the planet's spectral lines are used to determine the period. This last
is the most difficult to do because it means measuring the minute tilt of
the spectral lines, and from that tilt, the rotational period.
Some of the early attempts are listed in Table I and Table II.
Table I - early values for Uranus
Date Period Technique Person
It appears that people were closing in on a rotational period
between 10h and 1 lh, and this value showed up in textbooks of the time.
Actually the notion that Uranus has a rapid rotation goes back to Herschel
who thought he saw a polar flattening of the planet. v" Somewhat later
Laplace provided further qualitative support for HerscheTs deduction by
noting that the observed co-planar nature of the satellite orbits implied
that Uranus needed a substantial equatorial bulge to counteract the
disruptive perturbations of the Sun. However, the moon Miranda (smallest
and innermost) is on an orbit substantially inclined to the common plane
of the remaining satellites. Curious. The first substantial datum on
Uranus’s rotation was provided by Deslandres in 1902, who detected the
tilt, induced by rotation, of reflected Fraunhofer lines in the planet's
spectrum, thus proving the retrograde sense of the planet’s spin.
Table II - early values for Neptune
Date Period Technique Person
There was considerable scatter in the proposed period for Neptune.
It is farther from Earth than Uranus, so it is more difficult to observe. Even
the sense of its spin was uncertain (prograde or retrograde). It was often
supposed that the planet probably rotated in a retrograde sense. Finally,
Washington Academy of Sciences
63
since Moore and Menzel’s work on Neptune was unique, a re-examination
of the sense of spin was worthwhile. The sense of the spin is a crucial
factor in understanding the evolution of Triton’s retrograde orbit.
In the mid 1970’s Mike Belton and Sethanne Howard (Hayes),
who both worked at Kitt Peak National Observatory (in Tucson, Arizona),
decided to re-measure the rotational periods of Uranus and Neptune. The
Voyager mission was due to encounter Uranus in 1986 so they had to get
their data before Voyager got there so their work would help prepare the
Voyager mission for the Uranus encounter. They wanted to determine the
rotation period, sense of spin, and orientation of the spin axis. They
decided to use the spectra method for Uranus, and spectra and photometry
methods for Neptune.
This is their story of how this was done in the days before the
Internet, thumb drives, and laptops. The details of the math are omitted
for simplicity.
For the spectral work they chose reflected Fraunhofer lines. These
spectral lines come from sunlight reflected by the planet’s atmosphere. Of
course, planets do not shine on their own. They are seen by reflected
sunlight. The visible and near-infrared spectra of Uranus and Neptune
have strong Fraunhofer absorption lines making them good candidates for
the tilted spectral line approach.
No one had ever observed any regular variation in the light of
Uranus (Campbell’s work was questioned) so they did not use the
photometry method for Uranus. Actually Howard had done a small project
in the mid-1970s where several images of Uranus were co-added together
to increase the signal-to-noise. The result was a fairly featureless planet
with an increase in contrast in one hemisphere. Not believing the results,
the project was dropped. That was perhaps unfortunate because Voyager
later showed the same feature.
For the photometric work, they already knew that Neptune showed
variations in infrared light so they chose that spectral waveband for the
observations.
Gathering the Observations
Belton received observing time on the Kitt Peak 4 meter Mayall
telescope for this project. It was unusual to get 4 meter time for planetary
work. He used Kodak Illa-J plates to record the data/"1 Before there were
digital data, there were glass plates with a photographic emulsion
Summer 2015
64
embedded on them. The plate was baked in N2 for several hours to
increase its sensitivity. After cutting the plate to the proper size (in the
dark), one exposed the plate to the object of interest. The developed plate
looked like a negative, lighter where the spectral lines appeared (Figure
7).
Astronomical spectra typically look like a series of lines, some
wide some narrow. In this case, each strip of a line represents a chemical
element or molecule in the atmosphere of the planet. Figure 5 shows a
standard absorption spectrum spanning blue to red. The lines are not tilted.
Figure 6 shows a planetary spectrum with a tilted linelx.
400 500 600 700
I I I I
wavtlcnqth in runometcri (10 * m)
Figure 5 — standard absorption spectrum
Figure 6 — tilted lines. The top is coming towards the observer (blue shift), the
bottom is going away from the observer (red shift). The laboratory line (no tilt) is
shown in white in front of the tilted line.
Belton oriented the slit of the spectrograph so that it spanned the
planet from one side to the other. He took a timed exposure. Then he
would rotate the spectrograph slit by a few degrees and take another
spectrum. He was finished when he had rotated the slit all the way around
the planet. He obtained a set of nice spectral data from Uranus and
Neptune. Some spectra are shown in Figure 7 which shows three position
angles (angle of the slit on the planet) of Uranus and Neptune and the
lunar spectrum used to set the plate scale, x dispersion, Xl and intrinsic line
tilt. This is a developed plate (a negative of the original) so the lines are
not dark, they are light in hue.
Why do the spectral lines tilt? Uranus is rotating about its axis. At
any given time one side comes toward us, the other away from us. This
Washington Academy of Sciences
65
results in a classical Doppler shift of the light we see. Place the
spectrograph slit entirely across the planet. Then one end of the slit has
light receding from the observer (red shift). The other end of the slit has
light coming towards the observer (blue shift). A red shift will shift the
position of the spectral line to the right just a bit. A blue shift will shift the
position of the spectral line to the left just a bit. The amount of shift varies
with the location of the slit on the planet. Near the planet center there will
be no shift at all. The farther from the center, the greater the shift, hence
a tilt to the whole line as it covers the planet.
MOON
Figure 7 — spectra of Uranus, Neptune, and the Moon
Note that the widths of the spectra are different for the two planets.
That is because as seen from Earth Neptune is smaller than Uranus.
Figure 8 (upper portion) shows a schematic of a planet with the
spectrograph slit across it at approximately a 45° angle (this is called the
position angle). It took the powerful 4 meter telescope to do this because
the image of the planet had to be large enough to encompass the whole
Summer 2015
66
slit. Figure 8 (lower portion) shows how the placement of the slit connects
to the spectral line.
Belton developed the glass plates and handed them over to
Howard for reducing (i.e., obtaining the data).
Figure 8 — Image of Uranus with overlaid slit (top image). The parallel black lines
represent the slit. Tilted spectral lines (bottom image) are shown with two dotted lines
showing where the light from the planet appears on the spectral line.
Washington Academy of Sciences
67
The photometry method for Neptune was handled differently from
the spectral line method. The Kitt Peak 50" telescope had an infrared filter,
so the observer would see infrared light and not much else.x" The light
from Neptune passed through the telescope and filter to land on a
photomultiplier tube - turning photons into electrons. From there the
signal would appear on a Brown chart recorder (an antique observing tool)
which fed a continuous strip of paper through - similar to what happens
with a lie detector test. The strip of paper recorded the infrared signal from
Neptune for as long as one could observe through the night. If the signal
never changed as the night wore on, then there was little to see as Neptune
rotated. But if the signal dropped occasionally and in a regular manner
then the time between drops would give an estimate of the rotation period.
They hoped for the best and indeed found this particular signature for
Neptune. X1" However, the data were rough and not well defined.
Nevertheless, they agreed fairly well with the spectroscopic results.
Figure 9 shows a sample of the Neptune data. Time (in terms of fractional
periods) is plotted along the horizontal axis, brightness along the vertical
axis. Note the dip in the middle of the graph. They were seeing something
(unknown) that caused the light from the planet to decrease in a regular
way.
Figure 9 — infrared photometry of Neptune
Summer 2015
68
Data Reduction
Howard began reducing the spectral data. She chose to measure
O 0
orders 46 (near 5000 A ) and 47 (near 4900 A ). These lines were near the
center of the plate, and were of good exposure with no overlapping orders.
She identified individual lines using the lunar spectrum as a reference so
that she used only Fraunhofer lines. Step one was to “digitize” the data
with an automated scanning machine called a microdensitometer. The
exposed glass plate was placed on the platen and automatically moved
step by step to measure the density of each spot on the plate. She used a
20 x 20pm aperture stepping every 10pm both along and across the
spectra. Each order was digitized in overlapping strips 20pm wide and
10,240pm long. In this way each spot on the glass plate was turned into a
number stored on a 7-track magnetic tape. The microdensitometer was
controlled by a PDP 8 computing machine (way back there in early
computers). It took weeks of work just to get the numbers stored on a 7
track tape.xlv Today the data would be taken with a CCD (Charge Coupled
Device) chip and recorded digitally right at the start.
There were no floppy disks, DVDs, or thumb drives. This was long
before the days of the laptop; so she used a Vax minicomputer for the
actual data processing. Vaxen were very nice, robust machines.xv Kitt
Peak had one of these that everyone shared. A Vax can read a 7-track tape.
She wrote several computer programs to read the data from the tape. Each
microdensitometer scan was cross-correlated with a running set of
weights approximating the function that represented the slit. The position
of the line was defined as the zero-crossing of the numerical derivative of
the cross-correlation. The tilt angle was found by a least squares fit of a
straight line through the zero-crossings. In other words, she digitally
reproduced the tilted line. The results for two lines are shown in Figure
10. The left side shows a good fit. The right side shows a poor fit.
As a back-up she also made large hard copy prints of the plates
(aka Figure 7). She hand measured the tilts with a compass, protractor,
and ruler as a check on the automated procedure. Interestingly enough the
errors in the hand check were about the same as the errors with the
automated procedure.
Washington Academy of Sciences
69
Figure 10 — two spectral line fits
In the case of Uranus the root mean squared (rms)xvl deviation of
the line tilt angles was about ±1.7° and the standard error of the mean
approximately ±0.4°. The rms deviation for Neptune was much larger. It
was about ±3° with a standard error of the mean about ±0.9°.
Knowing the measured tilt angle is not enough. The line tilt must
be corrected for the astronomical seeing. This is the largest source of
possible error. “Seeing” is a measure of the quality of the sky. Is the image
a pinpoint or sharp (good seeing) or is it smeared out (poor seeing)?
Celestial objects blur and twinkle because of turbulent mixing in the
Earth’s atmosphere. Astronomers always hope for a clear night with good
seeing. It can happen that a night can be quite clear yet unusable because
of poor quality seeing.
The angular size of Uranus as seen from Earth is a known value.
One can then estimate how large the observed planet is with respect to
that known value. This is a measure of the seeing - in essence, how
“fuzzy” is Uranus. The same method is used for Neptune. Of course
“fuzzy” can be a bit subjective, hence the possibility of error. In this case
the value of the ‘seeing’ is the full width at half intensity of the Gaussian
smoothing function required to explain the distribution of density of the
plate across the dispersion (i.e., how tall is the spectrum). In other words,
use a Gaussian distribution to map the width of the planet (or height of the
spectrum).
From there one estimates the effective seeing corrections by
matching the cross-dispersion distribution of intensity on each plate with
Summer 2015
70
the intensity in a model spectrum that results from a convolution of a
Gaussian function and a model planetary limb darkeningxv" function at the
spectroscopic slit. That is a lot of fancy words that mean match the
observed profile of Uranus with the known profile of Uranus. The
difference between the two is a measure of the seeing. The estimate of
seeing therefore depends on the radius assumed for each planet. Howard
and Belton assumed Uranus to have a radius of 25,900 km (actual value
25,362 km) and Neptune to have a radius of 24,500 km (actual value
24,622 km).
For Uranus this correction meant the line tilt needed to increase by
32%. For Neptune it was a 202% increase (Neptune was small and fuzzy)!
How good was this estimate of the seeing? The uncertainty in the
corrected tilt arising from estimating the seeing was about ±8% for Uranus
and about ±12% for Neptune.
At long last they had the line tilt, 6. Onward to the rotation.
The angular velocityxvm, co \ is a measure of the rotation period, co
is directly related to the line tilt and is independent of the radius of the
planet.
There are a number of other things to consider when getting a
rotation rate, co \ from the line tilt. Ultimately the relationship between the
angular velocity, co. of a planet near opposition and the tilt of the spectral
line, tan 0. in reflected light is (see the original paper for derivation of this
equation):XIX
206265c £>(/ t)tan(9
(l + cos^sd X
where c is the velocity of light, D(k ) the plate dispersion, s the plate scale,
cf> the planetary phase angle, d the distance to the planet, i/a \ the position
angle of the slit, ^poiethe position angle of the pole, and f0 the latitude of
the observer (us) with respect to the planet’s equator. All the parameters
on the right hand side are known. Thus, one can solve for the parameters
on the left side: position angle of the pole, y/po ie, the rotation period, co,
and the spin direction (clockwise or counterclockwise). The position angle
is the angle measured in the plane of the sky going counterclockwise from
north. It is a standard tool in astronomy,
One can see that the actual data reduction is rather complex. After
all the work and calculations were done they had their answers.
(ycos/’@sin(^pole - y/s)
Washington Academy of Sciences
71
They found that the spin of Uranus is retrograde; the spin ot
Neptune is prograde. This confirmed early spectroscopic results.
For Uranus the direction of its pole points a little south of the
orbital plane, thus making Uranus truly a sideways planet. The position
angle of the pole, projected onto the plane of the sky, is 283° ±4 ( i.e ., 13°
south of the equator). As it revolves, Uranus rotates like a drunken
astronomer rolling around the floor. Near the time of Uranian solstices,
one pole continuously faces the Sun, and the other one faces away. Only
a narrow strip around the equator experiences a rapid day-night cycle with
the Sun low over the horizon as in Earth’s Polar Regions. Each pole gets
about 42 years of continuous sunlight, followed by 42 years of darkness.
Near the time of the equinoxes, the Sun faces the equator of Uranus giving
a period of day-night cycles similar to those seen on most of the other
planets.
For Neptune the pole points north of the orbital plane, agreeing
with earlier results. The value for Neptune is 32° ±11. Remember the
Earth’s pole is tilted about 23.5°.
The rotation periods Howard and Belton found differed
significantly from earlier work. For Uranus they obtained: 24 ± 3 hr. and
for Neptune: 22 ± 4 hr.
As an additional check they were fortunate to obtain from Lowell
Observatory the original plates taken in 1912 by Lowell and Slipher for
Uranus and Mars. They used the same data reduction method and found
significant differences between their Uranus/Mars data and Lowell and
Slipher’ s values. They have no explanation for these differences. They
also used spectra of Jupiter reduced the same way and found agreement
with the known rotation period.
Now done, they published the results. The work made a sizable
splash in the news, even showing up in the TVew York Times and Popular
Mechanics.
Naturally one rarely abandons a scientific project. A few years
(1980) later they discovered that their estimates of the seeing were in
error. The error had little effect on Uranus, but had a much greater effect
on Neptune due to its farther distance and smaller angular size.xx Basically
they learned that the limb darkening of Neptune may be about the same
as Uranus. They had assumed that the two planets had different limb
darkening values. Once they changed this parameter the corrected value
Summer 2015
72
for Neptune's rotation period became 15.4 ±3 hr. If the limb darkening is
less for Neptune, then the rotation period is lengthened.
Conclusion
What did Voyager find when it got there? For Uranus the rotation
period is 17h 14m. They were a bit off there. For Neptune the rotation
period is 16h 6m. So, oddly, they were closer for the more distant planet.
The axial tilt for Uranus is 97.77° (about 8° south of the orbital plane, they
got 13°). The axial tilt for Neptune is 28.32° - not too far from their
determination.
All in all, a real visit is worth the price of admission.
The two Voyager missions are still operating as they move through
the heliosheath - the place where the interstellar gas meets the solar wind.
They have long since left the Solar System planets behind. Their current
locations are continually updated on the Voyager website
http://voyager.jpl.nasa.gov/. Check it out. In 2015 Voyager had its 38lh
birthday, and it is the longest operating of any NASA satellite.
Uranus as seen by Voyager
Washington Academy of Sciences
73
Summer 2015
74
I Laplace (1749 - 1827) was a French mathematician and astronomer.
II English astronomer/mathematician.
III This effect had been suggested by the English astronomer Mary Somerville (1780 -
1872).
IV French mathematician.
v German astronomer.
V1 Galileo had probably observed Neptune, but he thought it was a star.
v" Basically, the flatter the planet the faster it rotates.
V1U Astronomers no longer use glass plates. Today everything is digital. But in the mid
1970’s glass plates were still common.
Ix The spectral lines from galaxies also tilt. One can use the line tilt to determine the
mass of the galaxy.
x The plate scale can be described as the number of degrees, or arcminutes or
arcseconds, corresponding to a number of inches, or centimeters, or millimeters
{etc.) at the focal plane (where an image of an object is “seen”).
XI The dependence of refraction on the wavelength of light is called dispersion. A lens
or prism disperses light.
XII Of course, no filter is perfect. They had to correct for leaks in the filter.
X1U When Voyager encountered Neptune it saw a “large spot” a storm rather like the
Great Red Spot on Jupiter. They must have been observing that spot as it rotated.
x,v People do not use 7-track tapes for data storage any more.
xv Vaxen are almost gone now too.
XV1 . A measure of the error.
xvu Limb darkening refers to the diminishing of intensity in the image of a star or planet
as one moves from the center of the image to the edge or “limb” of the image.
xvni Angular speed at which the planet is rotating.
xlx Howard-Hayes, S. and Belton, M., “The Rotational Periods of Uranus and Neptune”,
Icarus, 32, 383-401 (1977).
xx Belton, M. J. S, Wallace, L., Howard-Hayes, S., and Price, M. J., “Neptune’s
Rotation Period: a Correction and a Speculation on the Difference between
Photometric and Spectroscopic Results”, Icarus , 42, 71-78 (1980).
Bio
Sethanne Howard is an astronomer who has held positions with U.S
National Observatories, NASA, the National Science Foundation, and the
U.S. Navy. She was Chief of the U.S. Nautical Almanac Office, 2000-
2003. Her research specialty is galactic dynamics. She has also been active
in science education, especially concentrating on the history of women in
science.
Washington Academy of Sciences
75
Washington Academy of Sciences
1200 New York Avenue, NW
Room 1 1 3
Washington, DC 20005
Membership Application
Please fill in the blanks and send your application to the Washington
Academy of Sciences at the address above. We will contact you as soon
as your application has been reviewed by the Membership Committee.
Thank you for your interest in the Washington Academy of Sciences.
(Dr. Mrs. Mr. Ms.)
Business Address
Home Address
Email
Phone Cell
Preferred mailing address
Please indicate:
Type of membership
Business Home
Regular Student
Present Occupation or Professional Position
Please list memberships in scientific societies - and include office held:
Summer 2015
76
Journal of the Washington Academy of Sciences
Instructions to Authors
The Washington Academy of Sciences publishes its interdisciplinary
peer-reviewed Journal of the Washington Academy of Sciences four times
a year — Spring, Summer, Fall, and Winter.
1 . Deadlines for quarterly submissions are:
Spring - February 1 Fall - August 1
Summer - May 1 Winter - November 1
2. Draft manuscripts using a word processing program (such as
MSWord), not PDF.
3. Papers should be 6,000 words or fewer. With 7 or more graphics,
reduce the number of words by 500 for each graphic.
4. Include an abstract of 1 50-200 words.
5. Graphics must be in black & white or greytone. They must be
referenced in the text.
6. Use endnotes, not footnotes. The bibliography may be in a style
considered standard for the discipline or professional field
represented by the paper.
7. Submit papers as email attachments to the editor.
8. Include the author’s name, affiliation, and contact information —
including postal address. Membership in an Academy-affiliated
society may also be noted.
9. Manuscripts are peer reviewed and become the property of the
Washington Academy of Sciences.
1 0. There are no page charges.
Please see the Academy’s web site, www.washacadsci.org, for the library
subscription rate, listing of articles dating to 1899, and information on
accessing them.
Washington Academy of Sciences
Washington Academy of Sciences
Affiliated Institutions
National Institute for Standards & Technology (NIST)
Meadowlark Botanical Gardens
The John W. Kluge Center of the Library of Congress
Potomac Overlook Regional Park
Koshland Science Museum
American Registry of Pathology
Living Oceans Foundation
National Rural Electric Cooperative Association (NRECA)
Summer 2015
78
Delegates to the Washington Academy of Sciences
Representing Affiliated Scientific Societies
Acoustical Society of America, Washington Chapter
American Assoc, of Physics Teachers, Chesapeake Section
American Astronomical Society
American Ceramics Society
American Fisheries Society
American Institute of Aeronautics and Astronautics
American Meteorological Society, Washington, DC
Chapter
American Nuclear Society, Washington DC Section
American Phytopathological Society, Potomac Division
American Society for Cybernetics
American Society for Metals, Washington Chapter
American Society of Civil Engineers, National Capital
Section
American Society of Mechanical Engineers, Washington
Section
American Society of Microbiology, Washington Branch
American Society of Plant Biologists, Mid-Atlantic
Anthropological Society of Washington
ASM International
Association for Women in Science, DC Metropolitan
Chapter
Association for Computing Machinery, DC Area Chapter
Association for Science, Technology, and Innovation
Association of Information Technology Professionals
Biological Society of Washington
Botanical Society of Washington
Capital Area Food Protection Association
Chemical Society of Washington
District of Columbia Institute of Chemists
District of Columbia Psychological Association
Eastern Sociological Society
Electrochemical Society, National Capital Section
Entomological Society of Washington
Geological Society of Washington
Historical Society of Washington DC
Human Factors and Ergonomics Society, Potomac Chapter
Institute of Electrical and Electronics Engineers, Northern
Virginia Section
(continued on next page)
Paul Arveson
Frank R. Haig, S. J.
Sethanne Howard
Vacant
Lee Benaka
David W. Brandt
Vacant
Charles Martin
Vacant
Stuart Umpleby
Vacant
Vacant
Daniel J. Vavrick
Vacant
Mark Holland
Vacant
Toni Marechaux
Jodi Wesemann
Alan Ford
F. Douglas
Witherspoon
Chuck Lowe
Stephen Gardiner
Chris Puttock
Keith Lempel
Elise Ann Brown
Vacant
Tony Jimenez
Ronald W.
Mandersheid
Vacant
Vacant
Jeffrey B. Plescia
Jurate Landwehr
Vacant
Gerald P. Krueger
Murty Polavarapu
Washington Academy of Sciences
Delegates to the Washington Academy of Sciences
Representing Affiliated Scientific Societies
(continued from previous page)
Institute of Electrical and Electronics Engineers, Washington
Section
Institute of Food Technologies, Washington DC Section
Institute of Industrial Engineers, National Capital Chapter
International Association for Dental Research, American
Section
International Society for the Systems Sciences
International Society of Automation, Baltimore Washington
Section
Instrument Society of America
Marine Technology Society
Maryland Native Plant Society
Mathematical Association of America, Maryland-District of
Columbia- Virginia Section
Medical Society of the District of Columbia
National Capital Area Skeptics
National Capital Astronomers
National Geographic Society
Optical Society of America, National Capital Section
Pest Science Society of America
Philosophical Society of Washington
Society for Experimental Biology and Medicine
Society of American Foresters, National Capital Society
Society of American Military Engineers, Washington DC
Post
Society of Manufacturing Engineers, Washington DC
Chapter
Society of Mining, Metallurgy, and Exploration, Inc.,
Washington DC Section
Soil and Water Conservation Society, National Capital
Chapter
Technology Transfer Society, Washington Area Chapter
Virginia Native Plant Society, Potowmack Chapter
Washington DC Chapter of the Institute for Operations
Research and the Management Sciences (WINFORMS)
Washington Evolutionary Systems Society
Washington History of Science Club
Washington Paint Technology Group
Washington Society of Engineers
Washington Society for the History of Medicine
Washington Statistical Society
World Future Society, National Capital Region Chapter
Richard Hill
Vacant
Neal F. Schmeidler
J. Terrell Hoffeld
Vacant
Vacant
Hank Hegner
Jake Sobin
Vacant
D. S. Joseph
Vacant
Vacant
Jay H. Miller
Vacant
James Cole
Vacant
Eugenie Mielczarek
Vacant
Daina Apple
Vacant
Vacant
E. Lee Bray
Terrell Erickson
Richard Leshuk
Vacant
Russell Wooten
Vacant
Albert G. Gluckman
Vacant
Alvin Reiner
Alain Touwaide
Michael P. Cohen
Jim Honig
Washington Academy of Sciences
Room 113
1200 New York Ave. NW
Washington, DC 20005
NONPROFIT ORG
US POSTAGE PAID
MERRIFIELDVA 22081
PERMIT# 888
Return Postage Guaranteed
Was
03*2-
Volume 101
Number 3
Fall 2015
Journal of the
WASHINGTON
ACADEMY OF SCIENCES
Board of Discipline Editors iii
Introduction K. Borne iv
Big Data Analytics and Workforce Issues C. McNeely. 1
Big Data Adoption in the Health Care Domain E. Kuiler. 1 1
Big Data: Who's Accountable? J. Halm 23
Exploring Bias and Error in Big Data Research K. Seely-Gant, L. Frehill 29
Educating Data Scientists H. Topi, M. L. Markus 39
Everything Old is New Again L. Frehill. 49
Social Media Analysis for Higher Education .A. Berea et ai 63
Privacy in a Networked World H. Xu, H. Jia 73
Membership Application 85
Instruction to Authors 86
Affiliated Institutions 87
Affiliated Societies and Delegates 88
ISSN 0043-0439
Issued Quarterly at Washington DC
MCZ
LIBRARY
HARVARD
UNIVERSITY
Washington Academy of Sciences
Founded in 1898
Board of Managers
Elected Officers
President
Mina Izadjoo
President Elect
Mike Coble
Treasurer
Ronald Hietala
Secretary
John Kaufhold
Vice President, Administration
Vice President, Membership
Sue Cross
Vice President, Junior Academy
Vice President, Affiliated Societies
Gene Williams
Members at Large
Paul Arveson
Michael Cohen
Frank Haig, S.J.
Neal Schmeidler
Mary Snieckus
Past President
Terrell Erickson
Affiliated Society Delegates
Shown on back cover
Editor of the Journal
Sethanne Howard
Journal of the Washington Academy of
■ScvewcesOSSN 0043-0439)
Published by the Washington Academy of
Sciences
email: journal@washacadsci.org
website: www.washacadsci.org
The Journal of the Washington Academy
of Sciences
The Journal is the official organ of the
Academy. It publishes articles on science
policy, the history of science, critical reviews,
original science research, proceedings of
scholarly meetings of its Affiliated Societies,
and other items of interest to its members. It
is published quarterly. The last issue of the
year contains a directory of the current
membership of the Academy.
Subscription Rates
Members, fellows, and life members in good
standing receive the Journal free of charge.
Subscriptions are available on a calendar year
basis, payable in advance. Payment must be
made in US currency at the following rates.
US and Canada $30.00
Other Countries $35.00
Single Copies (when available) $15.00
Claims for Missing Issues
Claims must be received within 65 days of
mailing. Claims will not be allowed if non-
delivery was the result of failure to notify the
Academy of a change of address.
Notification of Change of Address
Address changes should be sent promptly to
the Academy Office. Notification should
contain both old and new addresses and zip
codes.
POSTMASTER:
Send address changes to WAS, Rm 1 1 3,
1200 New York Ave. NW
Washington, DC 20005
Academy Office
Washington Academy of Sciences
Room 113
1200 New York Ave. NW
Washington, DC 20005
Phone:(202) 326-8975
1 200 New York Ave
Suite 11 3
Washington DC
20005
www. washacadsci.org
Journal of the
WASHINGTON ACADEMY OF SCIENCES
Volume 101 Number 3 Fall 2015
Contents
Editorial Remarks S. Howard ii
Board of Discipline Editors iii
Introduction K. Borne iv
Big Data Analytics and Workforce Issues C. McNeely 1
Big Data Adoption in the Health Care Domain E. Kuiler 1 1
Big Data: Who’s Accountable? J. Hahn 23
Exploring Bias and Error in Big Data Research K. Seely-Gant, L. Frehill 29
Educating Data Scientists H. Topi, M. L. Markus 39
Everything Old is New Again L. Frehill 49
Social Media Analysis for Higher Education A. Berea et al 63
Privacy in a Networked World H. Xu, H. Jia 73
Membership Application 85
Instructions to Authors 86
Affiliated Institutions 87
Affiliated Societies and Delegates 88
ISSN 0043-0439 Issued Quarterly at Washington DC
Fall 2015
11
Editorial Remarks
Big Data is a buzzword we hear often these days. Datasets can be so large
or so complex that traditional data processing applications are inadequate.
We shall peek into a conversation that speaks to big data validity,
credibility, applicability, and its broader implications.
The Fall issue of the Journal is dedicated to a Symposium entitled “Big
Data Analytics and Workforce Issues: Initiatives, Research, and
Challenges” which was part of the 7lh annual Dupont Summit held this past
December in Washington, DC, and sponsored by the Policy Studies
Organization. The objectives of the Symposium included articulating
critical research and policy questions on big data and identifying problems
that must be faced to answer them. Moreover, the presenters discussed the
explosion of big data in and across different contexts (academia, industry,
and government - the “triple helix”) and at different levels of analysis.
The organizer of the Symposium was Connie McNeely from George Mason
University. The moderator was Jong-on Hahm also from George Mason
University.
Panelists were Philip Bourne, National Institutes of Health; Heng Xu,
National Science Foundation; Erik Kuiler, Systems Made Simple, Inc.; Lisa
Frehill, Energetics Technology Center; Michelle Schwalbe, National
Research Council; and Laurie Schintler, George Mason University.
We are fortunate that Connie McNeely gathered papers from the panelists
for publication in the Journal. Her paper leads the discussion and introduces
the rest of the contributors. Enjoy an unusual and unique look at Big Data.
Kirk Borne starts us off describing the concept of Big Data and providing
an introduction to the papers.
Sethanne Howard
Editor
Washington Academy of Sciences
Journal of the Washington Academy of Sciences
Editor Sethanne Howard
sethanneh@msn.com
Board of Discipline Editors
The Journal of the Washington Academy of Sciences has a 12-member
Board of Discipline Editors representing many scientific and technical
fields. The members of the Board of Discipline Editors are affiliated with a
variety of scientific institutions in the Washington area and beyond —
government agencies such as the National Institute of Standards and
Technology (NIST); universities such as Georgetown; and professional
associations such as the Institute of Electrical and Electronics Engineers
(IEEE).
Anthropology Emanuela Appetiti
Astronomy Sethanne Howard
Biology/Biophysics Eugenie Mielczarek
Botany Mark Holland
Chemistry Deana Jaber
eappetiti@hotmail.com
sethanneh@msn.com
mielczar@physics.gmu.edu
maholland@salisbury.edu
djaber@marymount.edu
Environmental Natural
Sciences Terrell Erickson
Health Robin Stombler
History of Medicine Alain Touwaide
Operations Research Michael Katehakis
Physics Katharine Gebbie
Science Education Jim Egenrieder
Systems Science Elizabeth Corona
terrell.ericksonl@wdc.nsda.gov
rstombler@aubuiTistrat.com
atouwaide@hotmail.com
mnk@rci.rutgers.edu
katharine.gebbie@nist.gov
i im@deepwater.org
elizabethcorona@gmail.com
Fall 2015
IV
Introduction
Big Data Nation - Foundations, Applications, and Implications
Kirk Borne
Principal Data Scientist, Booz Allen Hamilton
Data is the new oil, the new natural resource, the new black, the new
bacon, and the new gold rush. Such statements have been made in one form
or another, and most have been labeled as “big data” hype. Nevertheless,
despite the hype, the growth in data is unmistakably a real (and really big)
phenomenon. Fortunately, the growth is not just in the volume of our data
collections, but also in the value, opportunities, and insights that
organizations can now achieve through the exploration and exploitation of
their massive (and growing) data assets. The papers published here cover
several dimensions of this data-driven revolution in the business of
everything: business, education, research, government, finance, healthcare,
natural resources, our personal and social lives, and more. In the paper by
Topi and Markus (“Educating Data Scientists in the Broader Implications
of their Work”), the authors categorize data science into three bodies of
knowledge: Applications, Infrastructure, and Implications. If we re-label
the second of these as “Foundations”, then we have not only a useful
mnemonic (i.e., the three "-ations"), but we also have a sensible
categorization of the papers that are presented here.
The foundations upon which we build big data and data science
applications for discovery, insights, and innovation include basic research
and engineering. That includes academic research as well as data
engineering for infrastructure research and development. The paper “Social
Media Analysis for Higher Education” by Berea, Rand, Wittmer, and Wall
is an excellent example of academic research on mining a particular type of
data: social media text data. The authors explore students’ views of their
higher education experience through a common social media analytics
technique, sentiment analysis. The paper “Big Data: Who’s Accountable?”
by Hahm takes us through three case studies where bias and mis-
categorization of data have led to inaccurate (and sometimes controversial)
results - the importance of starting with a proper foundation in data
sampling, data integration, and interpretation of data analytics conclusions
is emphasized throughout. A related paper “Exploring Bias and Error in Big
Washington Academy of Sciences
V
Data Research” by Seely-Gant and Frehill examines research ethics
(foundations and implications) of sample bias and erroneous interpretations,
which are increasingly common in the big data era (especially when
working with social media, open source platforms, and online user data).
The applications of big data and data science are everywhere, and
there are papers here that examine those cases. The paper “Big Data
Adoption in the Health Care Domain: Challenges and Perspectives” by
Kuiler looks at applications in healthcare (improving patient care and
population health), while also addressing foundational issues (workforce
development) and implications (data anonymization and data privacy
confront data sharing). In the paper “Everything Old is New Again: The Big
Data Workforce”, Frehill looks at the abundant, novel, and game-changing
applications of big data across all sectors ( e.g ., business, health, and
finance). The author then explores how this changes workforce
development, education programs, consumer/customer experience, and the
landscape of data-driven decision-making in organizations.
Finally, the implications of what we are doing with our data
collections deserve special attention, both here as well as in data science
and analytics education programs. One of those areas of focus in academic
programs should be data ethics. There are very few examples of ethics
courses that specifically address the big data era - one of those is the “Data
Ethics in an Information Society” course in the George Mason University
Computational and Data Sciences degree program. Several papers here
address such societal and workforce implications. The aforementioned
paper by Topi and Markus specifically examines the legal, ethical, and
societal implications of analytics and data science, specifically in the
context of training data scientists and analysts to be aware of and diligent in
minimizing the possible harmful consequences of their analytics
applications. The paper “Big Data Analytics and Workforce Issues:
Prospects and Challenges in the Information Society” by McNeely
examines both applications and implications of big data analytics, with a
central focus on the latter, specifically: challenges across technical, social,
political, and economic dimensions. In the insightful paper “Privacy in a
Networked World: New Challenges and Opportunities for Privacy
Research” by Xu and Jia, the authors investigate new concepts,
consequences, and concerns related to privacy in our increasingly digital
Fall 2015
VI
lives. They examine human-data interaction, information linkability,
information ephemerality, and information identifiability. All of this
naturally leads to concerns about information liability, something which all
data analytics professionals must weigh on the balance sheet of information
and data assets.
As data emerges from the era of big data hype, contributing value
beyond our large data repositories, and blooms into a major organizational
asset and a major organizational product, the papers presented here will
provide valuable guidance, insights, and perspectives concerning the
foundations, applications, and implications of data analytics in a data-
drenched world.
Bio
Kirk Borne, PhD is an astrophysicist, Big Data science consultant, public
speaker, and the Principal Data Scientist in the Strategic Innovation
Group at Booz Allen Hamilton. He previously spent 12 years as tenured
Professor at George Mason University in the Computational Science,
Informatics, and Data Science programs. Before that, he worked 18 years
on various NASA contracts, as research scientist, as a manager on a large
science data systems contract, and as the Hubble Telescope Data Archive
Project Scientist. He also actively promotes data literacy by disseminating
information related to data science and analytics on social media, where he
has been named consistently since 2013 among the top worldwide
influences in big data and data science.
Washington Academy of Sciences
Big Data Analytics and Workforce Issues:
Prospects and Challenges in the Information Society
Connie L. McNeely
George Mason University
Abstract
Big data is one of the most critical features marking and defining our world
today. It constitutes an analytical space encompassing processes and
technologies that can be applied across a wide range of domains in the current
and growing information society. The articles presented in this issue address
related challenges and prospects as crucial considerations in technical, social,
political, and economic power and relations in national and international
contexts. With particular attention to conceptual delineations, analytical
applications, and educational and workforce dynamics, they attend to both
instrumental and intrinsic aspects of big data relative to society in general.
Together, the articles constitute a conversation that speaks to big data validity,
credibility, applicability, and broader societal implications — both positive and
negative — today and in the future.
Introduction
Big data, in all of its manifestations and applications, is the beating heart
of today’s burgeoning information society. On the one hand, big data has
been acclaimed in line with promises for societal benefits. However, on the
other hand, big data also has sparked controversies and debates on the
challenges and vulnerabilities that it has created relative to social, political,
and economic power and relations. Whether addressed in terms of technical,
social, or organizational perspectives, relevant topics are in the forefront of
initiatives in academia, government, and industry. Moreover, the advent of
big data has raised questions about those who use it and those who work
with it, especially in light of socio-cultural and structural dynamics and
disparities in terms of educational and workforce dynamics. Accordingly,
the objectives of the articles in this collection include articulating critical
research and policy questions and identifying challenges to answering them
and to engaging big data effectively. Employing perspectives that address
both instrumental and intrinsic aspects of big data, they offer an effective
and comprehensive view on the prospects and challenges of big data
analytics and workforce issues in and across different contexts and at
different levels of analysis. Together, the articles constitute a conversation
Fall 2015
2
that speaks to big data validity, credibility, applicability, and broader
societal implications now and in the future.
New opportunities and prospects, but also new challenges,
controversies, and vulnerabilities, have marked the explosion of big data as
a phenomenon in and of itself. In 2000, only a quarter of all stored
information was digital; by 2013, more than 98 percent of the world’s stored
information was digital (Mayer-Schonberger and Cukier 2013). Indeed,
“the world contains an unimaginably vast amount of digital information
which is getting ever vaster ever more rapidly. This makes it possible to do
many things that previously could not be done: spot business trends, prevent
diseases, combat crime, and so on. Managed well, the data can be used to
unlock new sources of economic value, provide fresh insights into science,
and hold governments to account” (Economist 2010) — but they also create
a host of new problems, with misuse and misinformation, security concerns,
privacy violations, etc. at the top of many related policy agendas. The ever-
increasing body of data is a core operational feature in virtually every sector
of society, and how we understand and use big data is increasingly the
defining feature of our times.
Conceptual Dimensions
Big data is a multidimensional concept referring to the exponential
growth and availability of both structured and unstructured data (SAS
2013), embracing technology, decision making, and policy. Big data has
largely been interpreted in terms of the “3 Vs”: volume, velocity, and
variety. That is, "big data is high volume, high velocity, and/or high variety
information assets that require new forms of processing to enable enhanced
decision making, insight discovery, and process optimization" (Beyer and
Laney 2012; Laney 2001). Volume indicates the increasing amount of data,
velocity indicates the speed of data, especially the rate at which it is created
or becomes available, and variety indicates the range of data types and
sources (Laney 2001). The compilation of large complex datasets has made
for massive volumes of data characterized by variety that reflect the
different types of structured and unstructured data that are collected;
velocity refers to how quickly these data can be made available for analysis
(UA 2015). Together, these dimensions comprise a basic model for
describing big data.
Washington Academy of Sciences
3
However, other “Vs” also have been included, especially variability
and veracity , such that reference to the “5 Vs” has become common. Along
with the variety and complexity that mark big data, variability is reflected
in inconsistencies in data flows (SAS 2013). The veracity of the data
represents an especially critical issue. Veracity is an indication of data
integrity and the extent to which it can be trusted for analytical and decision-
making purposes (UA 2015). Methods for data verification and validation,
as specifically applied to big data, are of particular importance in this
regard. In addition, another “V” — value — is sometimes discussed as a
separate dimension of big data, highlighting the value-added capacity of big
data (IDC 2012).
In any case, while there is a lack of consistent definition, the term “big
data” has reached some general agreement among various stakeholders as
constituting at least some indication of volume, signaling the size of datasets
as the critical factor (Ward and Barker 2013). After all, the allusion is to
“big” data in relative terms. Big data is derived from various sources, in
particular streaming data as the Internet of Things, social media data, and
publicly available open data. The conversion of large collections of
documents from print to digital format is giving rise to massive archives of
unstructured data, and social media, crowdsourcing platforms, and various
applications are producing reams of information from the real-time
transactions of people around the world. The complex structure, behavior,
and permutations of datasets are a fundamental consideration in describing
data as big (Ward and Barker 2013). However, having said that, “big data
is less about data that is big than it is about a capacity to search, aggregate,
and cross-reference large data sets” (boyd and Crawford 2012, p. 663).
Underlying the concept of big data are the technologies — the tools and
techniques — that are used to process massive or complex datasets. Hence,
we can refer to big data as a term describing the storage and analysis of
large and/or complex datasets using a series of applicable techniques (Ward
and Barker 2013).
From an expanded theoretical and practical perspective, big data also
has been described as a cultural, technological, and scholarly phenomenon,
resting on the interplay of technology, analysis, and mythology (boyd and
Crawford, p. 663):
Fall 2015
4
1) Technology : maximizing computation power and algorithmic
accuracy to gather, analyze, link, and compare large data sets.
2) Analysis : drawing on large data sets to identify patterns to make
economic, social, technical, and legal claims.
3) Mythology : the widespread belief that large data sets offer a
higher form of intelligence and knowledge that can generate
insights that were previously impossible, with the aura of truth,
objectivity, and accuracy.
In more encapsulated terms, big data reflects “a point of view, or
philosophy, about how decisions will be — and perhaps should be — made
in the future” (Lohr 2013).
Data-to-Knowledge-to-Action Analytics
“The challenge of big data is to convert it into useable information by
identifying patterns and deviations from those patterns” (UA 2015). In
epistemological terms, information is comprised by a collection of data, and
knowledge is established through different strands of information
(. Economist 2010), leading to questions that speak to the process of
converting data to knowledge to action. For example, what are analytical
and policy implications of the data in light of the how and why they are
collected, categorized, and aggregated? Do such data tasks reflect on how
they are or should be used? Furthermore, as elsewhere queried by McNeely
and Hahm (2014), will the analysis of big data provide insights and
information that will allow the development of answers to big questions, or
will it simply provide larger scale versions of answers already attained with
smaller data? Frankly, actual understanding is not stressed in most big data
approaches; correlations are the rule, representing a move away from
actually understanding phenomena to simply indicating associations
(Mayer-Schonberger and Cukier 2013).
The challenge of turning data into knowledge reflects matters of data
interpretation and re-purposing relative to secondary data markets
(Washington 2014). In practice big data might yield information, but not
necessarily understanding. Keep in mind that data gain meaning only in
context. What critical or fundamental factors must be considered for true
understanding? The socio-technical limitations of big data rest on
considerations of context and meaning and, as such, big data must be
engaged with an appreciation of both its power and its limitations. More to
Washington Academy of Sciences
5
the point, while big data is increasing, the ability to translate it into
knowledge and, more, to extract wisdom from it is relatively rare (McNeely
and Hahm 2014, p. 307; Economist 2010).
Large datasets have long been around and in use in various fields.
However, the big data revolution invokes a different frame for engaging
them. The integration of data from various sources and the use of that data
for purposes beyond those for which it was originally collected or created
are principal tasks associated with big data use (Berman 2013). Moreover,
at this point, it appears certain that data will continue to get “bigger and
bigger.” The Internet of Things is expected to comprise tens of billions of
objects by the end of this decade and is actively and instantaneously sensing
data on virtually every aspect of our lives and environment. Noting this
trend, Kuiler (p. 11) looks to the volumes of clinical, financial, and
consumer information available to healthcare organizations. Mapping a
complex multi-disciplinary approach to big data analytics, he focuses on
questions related to health and bioinformatics. In application, he categorizes
and reviews a wide range of structured and unstructured data and offers an
imiovative approach to performance measurement in the healthcare domain.
Overall, he provides evidence on the use of big data analytics for reducing
operational costs and optimizing performance, for improving regulatory
compliance, and for increasing returns on investments, while also
delineating future trends in big data analytics. However, he also explores
challenges and barriers to big data analytics and use, discussing limitations
and difficulties incurred in, for example, industry refusals to share data,
institutional barriers, and information governance. Framing big data as a
trope for a number of different technological and institutional factors, he
points to the problems of an abundance of data relative to a scarcity of
information, noting that more data is not always better.
Approaching such issues from a different direction, practical questions
of data veracity also are fundamental for converting data to knowledge to
action. Problems of sampling bias are particularly relevant in this regard.
Sampling bias is inherent in many big datasets. How might that affect policy
development and implementation? Seely-Gant and Frehill (p. 29) examine
related complications along these lines, discussing how sampling issues,
especially selection bias, associated with big data sources can have far
reaching implications for analysis and interpretation. Furthermore, in the
Fall 2015
6
same vein, Hahm (p. 23) offers a commentary on accountability and data
veracity, pointing out how sampling and sorting bias and errant
categorizations can lead to inaccurate conclusions, which can be
particularly dangerous for informing policy decisions.
While also encompassing these problems, one of the most prominent
and controversial issues that arises in discussions of big data is privacy. Xu
and Jia (p. 73) probe this topic, examining changing conceptions of privacy
in today’s big data environment in terms of information identifiability,
ephemerality, and linkability. They apply this conceptual approach to
investigate threats to information privacy in light of the collection and
analysis of large-scale data from social networking sites. Focusing on
human-data interaction, they turn to problems and risks associated with, for
example, data de-identification and re-identification, data integration, and
legal obligations and developments with regard to privacy issues. Their
primary emphasis is on mapping privacy regulations into actionable
information technology requirements that are re-usable across systems.
Education and Workforce Dynamics
Big data engagement and related topics are relevant within and across
sectors and require examination from technical, social, and organizational
perspectives. The skills, training, and education necessary for big data
related jobs in industry, government, and academia have become a focus of
discussions on educational attainment relative to workforce trajectories.
Especially given assertions of a skills gap for manipulating, analyzing, and
understanding big data, the relationship between education and the
development of the big data workforce is a critical point of departure for
delineating the field in general. Further, the role of big data in affecting
social, political, and economic relations and power come into play as
reflected in questions of educational and workforce opportunity and access,
and also raising questions of the “digital divide.” Do gatekeepers come into
play with big data, as in other fields, precluding certain individuals or
groups from accessing data or participating in relevant fields? In general,
basic questions on building proficiency in big data and workforce
development are at the forefront of debates in different sectors (NRC 2014).
That is, what should be taught, by whom, to whom, and how?
Washington Academy of Sciences
7
Technically speaking, training and education for big data jobs typically
require a basic knowledge of statistics, quantitative methods, or
programming, upon which applicable skillsets can be built. Such
background can be acquired in a number of fields that have long
incorporated related preparation. For example, Frehill (p. 49) notes that
“social scientists have worked with exceptionally large data sets for quite
some time, historically accessing remote space, writing code, analyzing
data, and then telling stories about human social behavior from these
complex sources.” However, she differentiates between traditional large
“designed” datasets and the new “organic” big data that are calling for more
and more trained knowledge workers with the required “deep skills and
talent.” Examining the role between higher education and the development
of the big data workforce, she addresses basic questions about participation
and also considers key lessons regarding gender differences in the big data
workforce.
Overarching changes in occupational roles and practices in the face of
technological shifts have led to revised workforce expectations and needs.
Some estimates suggest a shortage by 2018 of some 190,000 data scientists
in the United States, in addition to 1.5 million analysts and managers with
knowledge and skills to use analyses of big data to make effective decisions
(GovLab 2013; MGI 2011). Frankly, when an industry or field is growing
rapidly, “it is not unusual for a shortage of workers to occur until
educational institutions and training organizations build the capacity to
teach more individuals, and more people are attracted to the needed
occupations” (CEA 2014, p. 41). Thus, Topi and Markus (p. 39) investigate
the growing number of analytics and data science programs, arguing for the
need to include an emphasis on the implications and consequences of
practices and applications in related fields. They note the need for big data
workers “who are sensitive to data downsides as well as upsides” to achieve
the benefits of big data while avoiding harmful consequences. From yet a
different perspective, the investigation presented by Berea, Rand, Wittmer,
and Wall (p. 63) rests on social media analysis, using big data itself in their
research on big data analytics within education and related policies and
reflecting the changing data landscape.
Fall 2015
8
Conclusion
The effects of big data are being felt everywhere. As an analytical
space, big data encompasses processes and technologies that can be applied
across a wide range of domains “from business to science, from government
to the arts” ( Economist 2010), with positive and negative implications
depending on perspective and application. As such,
Big Data triggers both utopian and dystopian rhetoric. On one hand,
Big Data is seen as a powerful tool to address various societal ills,
offering the potential of new insights into areas as diverse as cancer
research, terrorism, and climate change. On the other, Big Data is seen
as a troubling manifestation of Big Brother, enabling invasions of
privacy, decreased civil freedoms, and increased state and corporate
control. As with all socio-technical phenomena, the currents of hope
and fear often obscure the more nuanced and subtle shifts that are
underway, (boyd and Crawford 2012, pp. 663-664)
Big data raises new issues and concerns related to, for example, privacy,
liability, security, and access, and has been invoked relative to new ways of
thinking about the world and relations across contexts. It has led to new
possibilities and prospects for research and policy, with fundamental issues
turning on cultural, organizational, and technological capacities at the heart
of debates and practices within and across academia, industry, and
government. Attending to issues of research and knowledge production, of
education and workforce dynamics, of socio-cultural, political, and
economic relations, the articles presented in this issue interrogate and
examine critical related issues from various perspectives, addressing
challenges and prospects for big data in theory and application in the
growing information society.
Washington Academy of Sciences
9
References
Berman, J.J. 2013. Principles of Big Data: Preparing, Sharing, and Analyzing Complex
Information. Burlington, MA: Morgan Kaufman.
Beyer, M.A., and D. Laney. 2012. “The Importance of ‘Big Data’: A Definition.”
Stamford, CT: Gartner.
boyd, D., and K. Crawford. 2012. “Critical Questions for Big Data: Provocations for a
Cultural, Technological, and Scholarly Phenomenon.” Information,
Communication, and Society 15(5): 662-679.
Chmura Economics and Analytics (CEA). 2014. “Big Data and Analytics in Northern
Virginia and the Potomac Region.” Northern Virginia Technology Council.
[https://gwtoday.gwu.edu/sites/gwtoday.gwu.edu/files/downloads/BigData%20repo
rt%202014%20for%20Web.pdf (accessed 3 September 2015)]
Economist. 2010 (27 February). “Data Data Everywhere: A Special Report on Managing
Information.” [https://www.emc.com/collateral/analyst-reports/ar-the-economist-
data-data-everywhere.pdf (accessed 3 September 2015)]
Eliot, T.S. 1934. The Rock: A Pageant Play. New York: Harcourt, Brace, and Company.
GovLab. 2013. “The GovLab Index: The Data Universe.” [http://thegovlab.org/govlab-
index-the-digital-universe (accessed 3 September 2015)]
IDC. 2012. “Worldwide Big Data Technology and Services, 2012-2015 Forecast.” IDC
Market Analysis #233485.
Laney, D. 2001. “3D Data Management: Controlling Data Volume, Velocity, and
Variety.” [http://blogs.gartner.com/doug-laney/files/2012/01/ad949-3D-Data-
Management-Controlling-Data-Volume-Velocity-and-Variety.pdf (accessed 3
September 2015)]
Lohr, S. 2013 (19 June). “Sizing Up Big Data, Broadening Beyond the Internet.” New
York Times, [http://bits.blogs.nytimes.com/2013/06/19/sizing-up-big-data-
broadening-beyond-the-intemet/?_r=0 (accessed 3 September 2015)]
Mayer-Schonberger, V., and K. Cukier. 2013. Big Data: A Revolution That Will
Transform How We Live, Work, and Think. New York: Houghton Mifflin Harcourt.
McKinsey Global Institute (MGI). 2011. Big Data: The Next Frontier for Innovation,
Competition, and Productivity.” www.mckinsey.com/mgi. (Accessed 3 September,
2015)
McNeely, C.L., and J. Hahm. 2014. “The Big (Data) Bang: Policy, Prospects, and
Challenges.” Review of Policy Research 3 1 (4): 304-3 1 0.
National Research Council (NRC), U.S. 2014. Training Students to Extract Value from
Big Data. Washington, DC: National Academies Press.
Fall 2015
10
SAS. 2013. “What is Big Data?” [http://www.sas.com/en_be/insights/big-data/what-is-
big-data.html (accessed 3 September 2015)]
University Alliance (UA). 2015. “What is Big Data?”
[http://www.villanovau.eom/resources/bi/what-is-big-data/#.VjBFMSuK820
(accessed 3 September 2015)]
Ward, J.S., and A. Barker. 2013. “Undefined by Data: A Survey of Big Data
Definitions.” [http://arxiv.org/pdf/1309.5821vl.pdf (accessed 3 September 2015)]
Washington, A. 2014. “Government Information Policy in the Era of Big Data.” Review
of Policy Research 3 1(4): 319-325.
BIO
Connie L. McNeely is a Professor in the School of Policy, Government,
and International Affairs at George Mason University, where she is also the
Co-Director of the Center for Science and Technology Policy. Her teaching
and research address various aspects of science, technology, and innovation,
healthcare, organizational behavior, public policy, governance, social
theory, and culture. Dr. McNeely directs major projects on big data
analytics, on scientific networks, and on migration and diversity in the
science and technology workforce, and leads an International Research
Group on Global Innovation in Science and Technology.
Washington Academy of Sciences
Big Data Adoption in the Health Care Domain: Challenges
and Perspectives
Erik W. Kuiler
George Mason University
Abstract
Due to recent technological advancements, health care organizations now have
access to large volumes of clinical, financial, and consumer information from
which to identify patterns and trends. As with other industries, health care is
grappling with the best ways to decipher and leverage these big data sets, with the
ultimate goals to enhance patient care and improve population health. The sheer
magnitude of the number of available data is both a boon and a hurdle. When
interpreting data, more information is not always better, unless an organization
assesses these data to discern what are noise and what are not. This paper explores
a number of challenges and barriers to big data analytics and use.
Introduction
The healthcare domain has witnessed a rapid growth in the delivery of
data-driven medicine resulting from the introduction of, for example,
electronic health records, digital imaging, digitized procedures, increasing
sophistication in lab test formulation, the real-time availability of sensor
data, and, what stands out in the popular press, the introduction of
genomics-related projects (Ohno-Machado 2012; Shah and Tenenbaum
2012). Information technology (IT) advances have led to a discourse on the
applicability of big data (a term coined by the Gartner Group, an IT industry
market research organization) to health data analytics (for example, Sahoo
et al. 2013). This study summarizes a presentation made at the 2014 Dupont
Summit, held in Washington DC, and explores topics considered in that
discussion. The paper concludes with a preliminary assessment of future
trends.
Adopting the Gartner Group’s definition, trade journals tend to
emphasize three big data properties, collectively referenced as the three V’s:
volume - to denote an exponentially large data set, ranging in size from one
or more terabytes ( 1 0 1 2) to multiple petabytes ( 1 0 1 5) or exabytes ( 1 0 1 8);
velocity - to indicate data that arrive as continuous streams, rather than as
transaction or database files; and variety- to designate data sets that contain
both structured and unstructured data that may be subject to different
Fall 2015
12
semantics and in different formats, gathered from diverse sources.1
Discussing big data in its historical perspective, Jacobs (2009) offers a
definition of the term that is perhaps more useful because it places big data
in its proper IT context: “Big data should be defined at any point in time as
"data whose size forces us to look beyond tried-and-true methods [of storage
and manipulation] that are prevalent at that time.,,, From Jacobs’ point of
view, in the 1960’s data files that could not be managed effectively with a
single tape mount could be considered as the big data of that era. Currently,
the capabilities to ingest, analyse, and manage multi-petabyte data sets have
underscored the limitations of our data analytics capabilities supported by
Relational DataBase Management Systems. These data management
limitations have led to the introduction of specific IT applications that
address the volume and velocity requirements of big data, much as in the
1960-80’s the availability of multi-tape data sets led to the introduction of
mechanical “tape monkeys” ( Jacobs’ term) to swap tapes in and out.
Because of its use in the popular press, the term big data has become
a trope for a number of different technologies and institutional conflicts
between the rights of the states and the rights of the citizenry: cloud-based
data and information analytics and big data management systems, data
interoperability as well as NS A spying, insurance denials based on big data-
based trend analyses, and security lapses that may lead to data breeches and
the loss of personally identifiable information (PII) - any data that,
collectively or severally, may potentially identify a specific individual
human being.
The scope of the healthcare domain is extensive, comprising the
activities of diverse epistemic communities, each of which has its own
institutional paradigms and cultural imperatives, resulting in a contested
equilibrium (adapting Amartya Sen’s phrase 1982, 1999) between different
interest groups: clinical health, focused on the delivery of patient-centered
healthcare services; public health, including clinical case surveillance,
syndromic surveillance, prevention, preparedness, and health promotion in
a community; population health, focused on health outcomes of a group of
individuals, including the distribution of such outcomes, in a population;
environmental health, focused on physical, chemical, and biological factors
external to a person, and all the related factors that may have an impact on
Washington Academy of Sciences
13
individual behavior; and, since the 1990’s, genomics - genes, genomes,
proteins, cells, ecological systems.
Impetus to Big Data Adoption in the US
In the United States, the impetus for big data-based health
informatics came during 2008-2010. Under the Health Information
Technology (HIT) for Economic and Clinical Health (HITECH) component
of the American Recovery and Reinvestment Act of 2009, the Centers for
Medicare and Medicaid reimburse health service providers for using
electronic documents in formats certified to comply with HITECH’s
Meaningful Use (MU) standards. The Patient Protection and Affordable
Care Act of 2010 (ACA) promotes access to health care and greater use of
electronically transmitted documentation. Health informatics are expected
to provide a framework for the electronic exchange of health information
that complies with all legal requirements and standards and, consequently,
expands the delivery of comparative effective- and evidence-based
medicine. HITECH MU and ACA support the adoption of Electronic Health
Records (EHR) as the preferred method for data interoperability among
patients, healthcare providers, and healthcare payers (HHS 2015).
Benefits of Big Data Adoption
Today, the availability of large data sets is the norm rather than the
exception. With the adoption of data analytics, end-users have become
increasingly more data literate, so that, while a simple spreadsheet would
have sufficed earlier, now end-users expect to see more complex models,
such as the results of time-series probability-based analyses to complement
snap-shot descriptive statistics.
The adoption of big data acquisition, management, and analytics
provides a number of important benefits: large sample size - the larger the
size, the greater the probability that the sample will accurately reflect the
characteristics of the population; increased predictive power - studies based
on big data samples are more likely to give statistically significant results;
a strong foundation for puiposeful action - cluster and category analytics of
very large data samples support the development of treatments and
protocols that are more accurately tailored to the specific needs of patient
populations (or cohorts).
Fall 2015
14
In addition big data analytics support proactive wellness and disease
management by discovering patterns; for example, snap-shots of current
operations, likely future trends, metrics of program efficacy and efficiency,
prospective needs of a population; decision support data to chart future
directions, data to support knowledge-intensive problem definition and
resolution (diagnostics, research, policy analysis, etc.). Also, big data
analytics enable healthcare improvements by, for example, integrating
clinical and claims data so that they are accessible, searchable, and
reportable; aggregating data from patient encounters to support public and
population health management; identifying and targeting individual patients
and cohorts for outreach; assessing quality of care across provider networks;
and correlating clinical and financial risk measures to optimize health care
delivery. Additionally, big data provide answers to important questions,
such as: How effective is a particular program, in terms of access and
results; is a client population served as well as it could be? Looking at
specific parameters, what policy changes should be enacted to make the
program better? Should more resources be allocated and of what kind,
when, and where? What is the likelihood of a patient suffering a stroke,
given his or her lifestyle, and, based on these probabilities, what kind of
ameliorative regimen should be proposed? What is the likelihood of a
provider committing fraud, given certain characteristics, and, as a corollary,
given the historical pattern of this provider’s behavior compared to that of
other providers, is this one committing fraud? How can a fraud detection
program be improved by operationalizing the analytics model?
Barriers to Big Data Adoption
While big data analytics can improve the delivery and quality of
healthcare, there are institutional barriers to their adoption, including
adversarial relationships between healthcare practitioners and HIT vendors,
lack of government incentives, economic limitations, and ethical and moral
constraints centered on data ownership, stewardship, and human rights.
Information Governance and Management Challenges
Prior to HITECH, health data sharing was usually limited to patient-
physician communications, and data interoperability between healthcare
providers was limited to facsimile distribution and similar dissemination
methods. The Internet changed all this, and the frequently cloud-based
Washington Academy of Sciences
15
aggregation of EHR data in very large collections (petabyte data sets, for
example), sustained by big data management and analytics, offers
opportunities to understand diagnoses, treatments, and protocols on a large
scale, providing an important complement to clinical trials. Nevertheless,
big data’s promise of increased data interoperability and information
sharing has exacerbated issues of syntactic conformity and semantic clarity
that have plagued data analytics since their inception in 1960’s automated
data processing environments. These issues require more than technological
solutions because the issues have their provenance in the cultural and
institutional determinants of the epistemic domains to which they apply,
rather than in the IT systems that support the analytics of such determinants.
Current HIT capabilities support the integration of data from diverse
sources that are frequently managed as data silos - for example, patient
clinical data, adverse event data, product data (drugs, medical devices,
blood, consumer, etc.), environmental and toxicological data, genomic data
industry-provided data (insurance, product, etc.) - without considerations
of data interoperability. Internally as well as externally, epistemic
communities in the health domain support different lexica and ontologies,
thereby restricting the possibility of efficient information sharing. For
example, in the clinical community, the International Statistical
Classification of Diseases and Related Health Problems Version 9 and 10
(ICD-9 and 10), which, although managed by the same agency, are not fully
compatible. Furthermore, these two standards are not fully compatible with
the Systematized Nomenclature of Medicine-Clinical Terms, another
frequently used standard, and require resource-intensive “cross-walks.” In
the genomics community, researchers employ at least two standards,
depending on where they operate in the world community: the GenBank file
format or the Swiss-Prott format. In the toxicology domain, the US National
Institute of Environmental Health Services’ participation in the
toxicogenomics ontology and global database initiatives is critically
important in establishing a common lexicon and ontology.
There are also different conveyance and transportation frameworks
for transporting data: the HL7 Version 2.x (V2) messaging standard is,
arguably, the most widely implemented standard for health data information
exchange in the world. However, this standard is not compatible with the
Fast Health Interoperable Resources Health Information Exchange and
Fall 2015
16
Clinical Document Architecture (CDA) specifications maintained by the
same organization. Moreover, neither the ICD-9/10 nor the HL7 standards
are compatible with the American National Standard X12 Electronic Data
Exchange transactions (the 274-278 and 834-835 series of transactions).
There are also different communication architectures in use: point-to-point
(peer-to-peer), and central repository (push/pull), etc.
Divergent Views of Product Requirements
Clinicians want HIT products that are tailored to support their
specific processes and protocols (based on my conversations with
practitioners and vendors at the 2014 AMI A national conference). Vendors
want to capture the largest market share possible at the lowest cost; hence,
the impetus to develop a generic, one-size-fits-all solution as the most
efficient model. There are also industry barriers to health data
interoperability. Many EHR vendors treat healthcare data as proprietary
assets that can offer considerable market advantages. Also, many lifestyle-
focused vendors (for example, in the tobacco, soft drink, and fast food
industries) resist health research and, consequently, data sharing.
Lack of Funding
The Federal government has offered programmatic incentives to
enhance healthcare delivery but these are frequently insufficient and not
sustained. For example, a number of programs, such as the Beacon
Community Cooperative Agreement Program, have come to an end. The
purpose of this program was to demonstrate how health IT investments and
the use of EHR’s could advance the vision of patient-centered care, while
achieving better health and better care at lower cost. The Health and Human
Services (HHS) Office of the National Coordinator for Health IT provided
$250 million over three years to 17 selected communities, each with its
unique population and regional context, throughout the United States that
had already made inroads in the development of secure, private, and
accurate systems of EHR adoption and health information exchange. When
the funding dried up, a number of Beacon Communities incorporated
elements of the program into their organizational structures and formed
consortia at their own costs but it is not likely, in the long run, that these
efforts can be sustained.
Washington Academy of Sciences
17
In 2014, 87% of US hospitals had some form of EHR system (Cohen
et al. 2014). In the US medical community many large institutions have
adopted the use of EHRs, data sharing, and big data analytics; however,
many small practices do not have the resources to adopt EHRs because they
are expensive and there are few incentives to support their adoption. Local
communities and regional governments usually do not have the resources
to assist medical practices with adopting EHR and big data analytics.
On the international level, many Southern Cone countries do not
have the resources to provide basic healthcare to their citizens, let alone
support an HIT infrastructure required to support EHR-based medicine, big
data management, and analytics (to which UN’s efforts to reach its
Millennium Goals can attest; see UN 2015). Likewise, International Non-
Governmental Organizations that have limited financial, organizational,
and temporal resources must frequently operate in adversarial environments
created by host governments.
Data Ownership and Data Stewardship
There are also institutional barriers to health big data analytics that
focus on data ownership and stewardship. Among the benefits of
introducing EHR’s and Personal Health Records (PHR) is to institutionalize
patient-focused healthcare so that patients become active partners in their
healthcare paradigms (for example, to mitigate patients’ strategic
ignorance: “my doctor knows what is best for me, and I expect her to notify
me when things may go wrong.”).2 Patients own their data; the medical
establishment and the government are data stewards. This uneasy alliance
raises questions of when, and under what circumstances these personal data
may be shared and how personal identity data can be protected against theft
and unauthorized access. To ensure the privacy of individually identifiable
health information in accordance with the Health Insurance Portability and
Accountability Act of 1996 (HIPAA), health data records must be
“anonymized” by removing all Personally Identifiable Information (PII)
from such records prior to their use in data analytics. Data anonymization
techniques are not fool-proof. A recent study noted that, in the absence of
PII, it is still possible to join records with a reasonable degree of accuracy
(for advertising purposes) from two discrete data sources based on date of
birth, 5-digit residential zip code, and gender (cited, among others, by
Fall 2015
18
Cavoukian and El Eman 2011; see also Kum et al. 2013 for an approach to
preserving privacy in interactive record linkages).
Legal , Moral, and Ethical Considerations
In the health domain big data analytics exacerbates the antinomy
between healthcare as a human right and healthcare as a commodity. The
UN Human Rights Charter and the Convention on the Elimination of All
Forms of Discrimination against Women formulate a concept of human
rights that includes rights essential to human development, such as rights to
adequate housing, healthcare, education, economic development (for
example, employment at a fair wage), that apply to all humanity, regardless
of gender, age, race, ethnicity, sexual orientation, etc. (UN 2002).
Nussbaum (2000; see also Sen 1982) observes that bodily health is second
only to life, supported by bodily integrity, as essential capabilities necessary
to flourish as individual human beings. Sen (1999) posits five types of
instrumental freedoms as essential to human freedom in the polity: political
freedoms, including free speech and free elections, to help promote
economic security; economic facilities, in the form of opportunities to
engage in market activities and production; social opportunities, among
them access to education and health care; transparency guarantees, those
mechanisms and institutions necessary to guarantee full disclosure of
information - the basis for trust; and protective security, those institutions
necessary to prevent any human from sinking into destitution and abject
poverty. Although there are moral strictures against using big data analytics
to restrict insurance coverage to individuals, such practices occur. Similarly,
in a “market model” of healthcare access and delivery, there are very few
means, other than moral approbation, to restrain a pharmacological drug
company from raising the price of a drug by, for example, 2000% or 5,500%
(CBS News September 22, 2015; NBC News September 22, 2015).
Big data analytics also faces other normative barriers. For example,
patients are likely to accept the necessity of data sharing among providers
to improve the quality of care, but the notion that their data will be shared
with other non-provider third parties has proven to be controversial,
especially when there are high-profile cases when data are shared without
the owner’s consent. The case of Hilda Lacks comes to mind. She was a
young African American woman who, in 1951, died of cervical cancer.
Doctors took samples of her cells without her knowledge and shared them
Washington Academy of Sciences
19
with other clinicians and researchers. Although labs were selling samples
of what came to be known as the HeLa cells, Lacks’s family received no
portion of the money generated by those sales and were not informed how
these cells were used (Falik 2014). Misuse of big data analytics and their
enabling technologies have fostered an increasingly greater wariness of
citizens of their government. The majority of citizens understand that their
data need to be shared to support the common weal but, as the activities of
the National Security Agency’s spying on the U.S. population indicate, the
citizenry is justified in its suspicions of its government.
Big Data Analytics Workforce Challenges
To be effective, big data analytics require a non-insular, non-
compartmentalized ontological perspective and a multi-disciplinary,
holistic approach to knowledge acquisition that incorporates skills from a
variety of academic disciplines, including, for example, quantitative
analysis (statisticians, computer scientists), finance (financial analysts, cost
analysts, fraud analysts), healthcare (medical practitioners, biologists,
chemists, product engineers), infrastructure and device engineering
(communications and device engineers), social sciences (sociologists,
medical healthcare economists), governance (policy analysts), information
management (information governance experts and managers, librarians),
deontology (ethicists), jurisprudence (legal professionals). The majority of
universities and training institutes do not offer cross-field programs that
emphasize the integration of these skills. As a result, one of the roles
frequently overlooked in efforts to minimize the risks of the misuse of big
data analytics is the role of the ethicist. If a little data analytics can lead to
misuse, big data analytics can lead to even greater misuse because so many
more data are available for abusive practices.
Future Trends
The study indicates a number of trends. Cloud-based big data
analytics and usage will grow. Big data analytics and data interoperability
have introduced increased concerns for effective privacy and security
management, defense against data breaches, and data storage management.
To address these concerns, government participation in developing,
promulgating, and enforcing standards will increase (for example, NIST
Standards for cloud-based security). Business intelligence (BI) vendors will
Fall 2015
20
expand their offerings to accommodate very large data sets and big data
analytics. Legal and human rights debates will become more contentious,
especially about topics such as data ownership and stewardship as they
relate to data sharing and individual privacy. In the health domain EHR
adoption will continue sporadically and in geographic isolation, especially
by the less-funded practices. Because of the extensive knowledge base
required to perform big data analytics effectively and ethically, big data
analytics will become increasingly the domain of intellectual and educated
elites.
References
Cavoukian, A., and K. El Eman. 2011. Dispelling the myths surrounding de-
identification: anonymization remains a strong tool for protecting privacy. Toronto,
Canada: Infonnation and Privacy Commissioner of Ontario.
Cohen, G., R. Amarasingham, A. Shah, B. Xie, and B. Lo. 2014. “The legal and ethical
concerns that arise from using complex predictive analytics in health care.” Health
Affairs, 33(7): 1139-1147.
CBS News. 2015. CEO: 5,000-percent drug price hike "not excessive at all."
[~http://www.cbsnews.com/news/turing-pharmaceuticals-ceo-martin-shkreli-defends-
5000-percent-price-hike-on-daraprim-drug/ (accessed 22 September 2015)].
Falik, D. 2014. “For Big Data, Big Questions Remain”. Health Affairs 33(7): 1111-1114.
Jacobs, A. 2009. “The pathologies of Big Data”. Communications of the ACM 52(8): 36-
44.
Kum, H., A. Krishnamurthy, A. Machanavajjhala, M. Reiter, and S. Ahalt. 2014.
“Privacy preserving interactive record linkage (PPIRL).” Journal of the American
Medical Informatics Association 21: 1-4.
Ohno-Machado, L. 2012. “Big science, big data, and the big role for biomedical
informatics.” Journal of the American Medical Informatics Association: 19: el.
NBC News. 2015. Price Fiike for Tuberculosis Drug Cycloserine Rolled Back From
2,000% Jump, [http://www.nbcnews.com/health/health-news/price-hike-
tuberculosis-drug-cycloserine-rolled-back-2-000-jump-n43 1716 (accessed 22
September 2015)].
Nussbaum, M. 2000. Women and human development: the capabilities approach.
Cambridge: Cambridge University Press.
Sahoo, S., C. Jayapandian, G. Garg, F. Kaffashi, S. Chung, and A. Bozorgi. 2014. “Heart
beats in the cloud: distributed analysis of electrophysiological ‘big data’ using cloud
Washington Academy of Sciences
21
computing for epilepsy clinical research.” Journal of the American Medical
Informatics Association 2 1 : 263-27 1 .
Shah, N. H., and J.D. Tenenbaum. 2012. “The coming of age of data-driven medicine:
translational bioinformatics’ next frontier.” Journal of the American Medical
Informatics Association 19: el-e2.
Sen A. 1982. Commodities and capabilities. Oxford University Press.
Sen, A. 1999. Development as freedom. New York: Knopf.
United Nations (UN). 2002. Human rights: a compilation of international instruments.
Vol. 1, Parts 1 and 2. New York: United Nations.
United Nations. (UN) 2015. Millennium development goals reports.
[http://www.un.org/millenniumgoals/reports.shtml (accessed 22 September 22
2015)].
United States Department of Health and Human Services (HHS). (2015). Beacon
Community Program, [http://www.healthit.gov/policy-researchers-
implementers/beacon-community-program (accessed 22 September 2015)].
United States Department of Health and Human Services (HHS). 2015. Health IT
Legislation and Regulations, [http://www.healthit.gov/policy-researchers-
implementers/health-it-legislation (accessed 22 September 2015)].
Endnotes
1 See, for example, periodic issues dedicated to the use of big data published in such
trade journals Federal Computing Weekly or Healthcare Informatics. Frequency of
webinars dedicated to big data offered by The Data Warehouse Institute (TWDI) may
also prove instructive.
2 With the advent of big data and the increasing of HTML-based EHRs (with the HL7
Consolidated Clinical (CCDA), it is possible to embed genomic data in a patient’s EHR,
offering the possibility of developing individualized medical protocols for patients.
Fall 2015
22
Bio
Erik W. Kuiler has spent most of his career as an Information Engineer,
focusing on the development of lexicons, ontologies, and systems to support
the management of data and information as enterprise assets. His research
interests include software and information engineering, rhetoric and
information theory, medieval studies and comparative literatures, and
public policy. His recent publications are on big data analytics, development
economics, and policy analysis.
Washington Academy of Sciences
23
Big Data: Who’s Accountable?
Jong-on Hahm
George Washington University
Abstract
Data analysis. Big or Small, requires careful handling of data to ensure against
sorting bias and errant categorization that can lead to inaccurate conclusions.
Sorting error may be introduced when attempting to hannonize existing datasets
with new datasets that offer many more parameters. Caution in data
categorization, searching for specific factors, and drawing conclusions, is
paramount for policymakers looking to use Big Data for societal benefits.
A state decides to SET aside economic development zones and wants to
encourage minority residents to establish businesses and hire workers. To
determine target areas for public outreach, data are scraped from publicly
available sources, cleaned, analyzed, and mapped to identify communities
where resources should be expended. When program administrators pull up
the first map, they are surprised to discover complete blanks in regions
where significant populations of minorities are known to reside.
A retailer planning an expansion into a new region pursues a
marketing scheme intended to identify specific populations. The first
advertising blitz includes a direct mail campaign where residents are offered
free trials and samples of products. After the first mailers are sent, the
retailer is bombarded with negative feedback and complaints about
inappropriate and offensive product information.
A consulting firm hired to improve efficiency and service in a
surgical unit at a hospital talks to every staff member of the surgical unit. It
devises an electronic tracking system to harmonize scheduling, smooth
patient transfer protocols and keep the unit at high functional capacity.
Within the first week, the schedule has bogged down completely and
patients have had to be referred to nearby hospitals.
In all of the above examples, something in the use of data has led to
unintended, sometimes risky outcomes. As has become clear, the use of Big
Data presents its own set of methodological and analytical challenges. The
question then arises: when Big Data goes wrong, who’s accountable?
Fall 2015
24
The advent of Big Data has seduced the unwary into the promise of
the possible, overlooking the promise of accompanying problems. In
enormous data sets, very slight differences in sorting and categorization can
lead to large differences, particularly as data are collected into ever-larger
sets. In a long-term study, when data increase by orders of magnitude over
time, minute differences could grow into significance and result in greatly
disparate impacts.
When examining accountability, the goal should not be to conduct a
forensic analysis of where, what, and how things went wrong, but rather to
establish parameters from the outset that prevent such errors in the first
place. In Big Data analyses, critical issues must be considered in the
research design: sorting bias and harmonization of old and new datasets.
Oftentimes, data collection is driven by the tools available. For
example, if a research project will use a certain analysis program or
approach, data will likely be collected in a format most amenable for use
with that program. If the data become unwieldy or the desired granularity is
different, sorting characteristics may be changed to make it work better with
the analytical tool. Sometimes, data may be sorted according to a
researcher’s own categorization algorithm without any conscious
realization of such sorting.
In 2013, the Wikipedia community noticed that its list of “American
novelists” no longer contained any women. They had all been placed on a
separate list of “American women novelists” (Filipacchi 2013). At the time,
Wikipedia wanted to make the list for “American novelists” less unwieldy,
and began creating subcategories (Neary 2013). This separation unleashed
a firestorm of criticism from those who perceived the categorization as
sexism, as there was no subcategory for “American men novelists”
(McDonough 2013; Flood 2013).
Whereas the constraints of Wikipedia’s platform may have led to
this reclassification, the bias in characterizing standard “American
novelists” to be male illustrates the potential for sorting bias in handling
data even before analysis is attempted. Data can be altered through
stratification, separation, or combination. Data cleaning can alter datasets
such that nuances of the raw data, small shifts, offsets or trends that may
indicate the influence of unanticipated factors is lost.
Washington Academy of Sciences
25
The negative impact of bias in data analysis and its impact have been
noted, sometimes in spectacular fashion. Photo auto-tagging programs
developed by Yahoo and Google led to African-Americans being tagged
with terms like “gorilla” and “ape” (Dwoskin 2015). Particularly in
software that uses machine learning algorithms, a biasing factor such as
selective data sets used for training can push the learning in an unintended
direction.
Unfortunately, avoiding bias is proving to be a much more daunting
task than previously conceived. According to Valerio Pascucci, a leading
researcher (XSEDE15, July 28, 2015), simply looking for something in a
large data set will inject bias into the data analysis (Gibson 2015).
Ideally, such unintended directional tacks will not occur in most
research projects. Data will be derived, for the most part, from known data
sets, analysis will use well-established, commonly used methods, data
comparability and compatibility will not be an issue. The challenge arises
when new data sets, offering richer, more informative views of a subject
population become available, and researchers want to incoiporate the new
information with the established set. At this point, the difficulty of data
harmonization, in even the most basic ways, becomes evident.
As computational capacity increases, an obvious difficulty arises in
matching categories end to end. The US Census in 2020 will include an
expanded range of choices for race and ethnicity to reflect the growing
percentage of Americans who identify as multiracial (Pew Research Center
2015).
In the example of the state wanting to spur minority
entrepreneurship and workforce development, the state government may
have wanted to harmonize existing data with newly available data on its
minority populations. The state may have had to rely on existing district
maps with much more limited population information. Constrained by
budget restrictions and the tools available, harmonization of data sets may
not have been as tailored as could be, leading to incorrect targeting of
desired populations. The retailer may have introduced sorting bias that
conflated demographics with product interest.
The third example is based on an actual case and illustrates how
sorting bias can creep into even Small Data analyses. The consultant hired
Fall 2015
26
to help improve surgery scheduling developed an electronic scheduling
system to replace a whiteboard on which surgery schedules were manually
written with markers. The system degenerated because the consultants did
not take into consideration the key role played by technicians who
maintained supply stocks in the surgical suites. The technicians were not
interviewed in the original data collection, nor were they provided access to
the new electronic scheduling system. Hence, they did not stock the suites
appropriately for each procedure. After a week of confusion and referral of
patients to nearby hospitals (with concomitant loss of revenue), surgical
scheduling reverted to the whiteboard system.
The very nature of Big Data underscores the potential impact of
infinitesimal differences that can become magnified in the petabytes of data
being generated and mined for meaningful information. While analytical
models can be changed with little imprint, impact on society and
communities lingers on.
References
Dwoskin, E. 2015. “How social bias creeps into Web technology.” The Wall Street
Journal, 21 August. http://www.wsi.com/articles/computers-are-showing-their-
biases-and-tech-finns-are-concerned- 1440 102894 (accessed September 23, 2015).
Filipacchi, A. 2013. “Wikipedia’s Sexism Toward Female Novelists.” The New York
Times, 24 April, http://www.nytimes.com/2013/04/28/opinion/sunday/wikipedias-
sexism-toward-female-novelists.html (accessed September 23, 2015)
Flood, A. 2013. “Wikipedia bumps women from ‘American novelists’ category.” The
Guardian, 23 April, http://www.theguardian.com/books/2013/apr/25/wikipedia-
women-american-novelists (accessed September 24, 2015)
Gibson, S. 2015. Exploring Farge Data for Scientific Discovery. HPCwire, 27 August.
http://www.hpcwire.com/2015/08/27/exploring-large-data-for-scientific-discovery/
(accessed September 23, 2015)
McDonough, K. 2013. American women novelists segregated by Wikipedia.
Salon.com, 25 April.
http://www.salon.com/2013/04/25/wikipedia moves women to american women
novelists category leaves men in american novelists/ (accessed September 24,
2015).
Washington Academy of Sciences
27
Neary, L. 2013. What’s in a Category? “Women Novelists” Sparks Wiki Controversy.
National Public Radio , 29 April, http://www.npr.org/20 13/04/29/1 79850435/what-s-
in-a-categorv-women-novelists-spark-wiki-controversv (accessed September 24,
2015).
Pew Research Center. 2015. “Race and Multiracial Americans in the U.S. Census.”
Multiracial in America, 11 June.
http://www.pewsocialtrends.orR/20 15/06/1 1 /chapter- 1 -race-and-multiracial-
americans-in-the-u-s-census/ (accessed September 28, 2015).
BIO
Jong-on Hahm is Special Advisor for International Research in the Office
of the Vice President for Research at George Washington University, and is
Distinguished Senior Fellow in the School of Policy, Government, and
International Affairs at George Mason University. Her research focuses on
global investments in science, technology, engineering, and mathematics
(STEM) to spur innovation and economic growth; global STEM workforce
migration; diversity in science; global intellectual property rights; and Big
Data uses, analyses, and controversies.
Fall 2015
28
Washington Academy of Sciences
29
Exploring Bias and Error in Big Data Research
Katie Seely-Gant and Lisa M. Frehill
Energetics Technology Center
Abstract
The availability and usability of massive data sets have added to the increasing
popularity of big data research. However, common mechanisms of big data
collection ( e.g ., social media, open source platforms, and other online user data)
can be problematic. Sampling issues, especially selection bias, associated with
these data sources can have far reaching implications for analysis and
interpretation. This paper examines the types of sampling issues that arise in big
data projects, how and why biases occur, and their implications. It concludes by
providing strategies for dealing with sampling and selection bias in big data
projects.
From the dawn of humankind to the year 2003, it is estimated that 5
exabytes (101S bytes) of data were created by humans. Today, humans create
about 2.5 exabytes of data every day (Intel IT Center 2012; Sagiroglu and
Sinanc 2013). This explosion of information, due in large part to
developments in data mining and collection, data warehousing and storage,
and computational capacity and performance, has led to “the era of big
data.” Individual-level data can now be collected and mined using online
platforms, social media, and cell phone applications, giving big data
researchers increasing levels of insight into previously unobserved
behavioral patterns and other “found data” (Harford 2014; Tufekci 2014;
Fan et al. 2014; Sagiroglu and Sinanc 2013; Yang and Wu 2013; De Mauro
et al. 2014).
With this influx of data researchers have been able to make significant
strides in fields such as health care, finance and economics, and social
science; however, big data research is not a panacea for data analytics.
Though some data scientists subscribe to the “myth of large n” - i.e., when
data are “big” enough biases are not significant -statistical errors and biases
can still impact research findings regardless of the size of the dataset
(Harford 2014; Anderson 2008; Lazer 2014). Due to the nature of data
collection and mining, and the methods used therein, big data research may
be particularly susceptible to sampling biases (Tufekci 2014; Fan et al.
2014; Harford 2014).
Fall 2015
30
This paper will explore the dangers in the “myth of large n” by
examining issues related to selection bias in big data research in particular,
and will attempt to assess the extent to which big data projects may be
affected by selection bias, the implications of this bias for research, and
potential strategies for accounting for such limitations. We choose to focus
on selection bias due to both the popularity of found data and mined social
media data in big data analyses and the likelihood that these sources produce
non-random samples.
Big data are characterized by the “3 V’s”, volume (amount and size),
velocity (real time or batch), and variety (structured or unstructured)
(Sagiroglu and S inane 2013; De Mauro et al. 2014). The source and utility
of big data can take many forms. Companies like Wal-Mart and Target
collect and analyze real-time purchase data to predict consumer preferences,
while economic researchers collect cell phone location data to determine
the distance consumers are willing to travel to a shopping mall, a proxy for
consumer demand and economic strength (Lazer 2014; Bollier 2010).
Health and life science researchers have harnessed big data and increased
computing power to revitalize genomic sequencing, compressing what had
been a ten-year process to less than a week (Fan et al. 20 1 4; Harford 20 1 4).
Covering all “3 V’s”, data mined from social media, cell phone
applications, and open source online platforms provide big data researchers
with unique insight into human behavior and interactions by providing
large, real time data on their users and their content (Bollier 2010). With
such large n’s, big data research offers interesting new ways to conduct
analyses. In lieu of the traditional method of formulating hypotheses and
theory before analyzing the data, big data researchers often take a high-level
look at massive data sets, noting interesting or unexpected correlations, and
then forming hypotheses and theories around those correlations (Lohr 2012;
Harford 2014; Anderson 2008; Bollier 2010). This exploratory approach
has led some to term the era of big data as the “end of theory” (Anderson
2008; Bollier 2010), suggesting that deductive reasoning grounded in
previous research literature is no longer necessary with such large, timely,
and varied data.
By pursuing these exploratory approaches, and making claims based on
discovered correlations, researchers risk falling prey to the traditional
limitations and biases inherent in both statistical and social research. When
Washington Academy of Sciences
31
researchers subscribe to the “myth of large n”, or the “n = all” suppositions,
certain faults of big data may be ignored (Lazer 20 1 4; Bollier 20 1 0; Harford
2014). In particular, big data are especially susceptible to endogeneity, auto-
correlation of errors, spurious correlations, and selection bias (Fan et al.
2014; Lazer 2014; Tufekci 2014; Harford 2014).
The emphasis on social media data mining and other data collection
from open source platforms and applications increases big data
vulnerability to selection bias in particular (Fan et al. 2014). Selection bias
describes the bias that is present when the selection of a sample or study
group is such that proper randomization is not achieved and the sample is,
therefore, not representative of the larger population (Berk 1983; Heckman
1979). Figure 1 shows a classic image that resulted from selection bias. The
Chicago Tribune , which relied heavily on telephone surveys for their
election predictions, prematurely claimed Dewey as the winner of the 1948
presidential election. The high cost of telephone lines led to biased results
as affluent Americans were more apt to support Dewey than those who were
less affluent. Selection bias is particularly problematic when relying on
exploratory research and analyzing correlations, since self-selected samples
often exhibit different correlational tendencies than random samples. Most
important for big data analysis is the presence of confounding variables, or
that persons who self-select into certain groups often have other variables
in common that researchers are not accounting for such as demographic
factors or similar environments, which cause the confounding variables
(Tufekci 2014; Fan et al. 2014).
More simply, selection bias describes the likelihood that certain persons
or groups are more apt to be picked up by big data collection efforts than
others, whether due to their use of social media and open source platforms,
the availability of internet connectivity in certain areas, their ability to
purchase smart phones and access applications, or any other number of
omitted variables. For example, StreetBump, a smart phone based
application rolled out in Boston, sought to record potholes while users were
driving and report these potholes to the city for repair. Eventually when
potholes were being disproportionately reported in affluent neighborhoods,
a deeper analysis revealed an issue with selection bias. Residents in affluent
neighborhoods were far more likely to own smart phones with network
Fall 2015
32
access — and cars — than residents who lived in lower income
neighborhoods (Harford 2014).
Figure 1. Classic Case of Selection Bias: Chicago Tribune Declares “Dewey Defeats
Truman” (photo credit: Associated Press).
These demographic findings are consistent with 2014 Pew Research
Center survey findings on smart phone users. Pew’s survey estimated that
about 65 percent of American adults are smart phone users, but that
population is, on average, under 40 years old, college-educated, and affluent
(i.e., average incomes of more than $75,000 per year), and far more likely
to be employed than non-smart phone users (Pew Research Center 2015).
Given the demographics of smart phone users, it is problematic to use smart-
phone data to make general claims about the U.S. population. Additionally,
these problems may be compounded if researchers take exploratory
approaches and simply analyze the data for interesting correlations, as
additional context is usually needed to determine what variables are driving
the correlation and what factors may cause the correlation to break down
(Harford 2014).
Selection bias has serious implications when left uncontrolled in a
standard linear model as it creates a non-linear relationship between the
dependent and key independent variable, such that a causal relationship may
be misinterpreted and effects resulting from random noise in the model are
mistaken for causal effects, affecting both internal and external validity.
Washington Academy of Sciences
33
External validity is undermined when selection bias is present by
underestimating the slope of the regression line, often leading to causal
effects being underestimated. Internal validity may be similarly
compromised if the effect of the “exogenous variable and the disturbance
term are confounded”, leading to causal effects of an independent variable
being confused with random noise in the data (Berk 1983, p. 388). By
failing to formulate a theory and corresponding model, researchers are often
unable to control for significant selection biases and jeopardize the validity
of their research (Berk 1983; Heckman 1979).
Other big data mining and collection efforts that have assumed “large n
= no bias,” analyze social media user data. For example, Twitter-generated
data has become quite popular among big data researchers because of the
ease of data collection along with its connection to user data, such as tweet
content, retweets, and engagement in trending topics. However, only 23 of
U.S. adults already online use Twitter, and among that 23 percent, the
population is largely minority (African-American and Hispanic) youth, with
about 37% of Twitter users under the age of 30 (Duggan et al. 2015). Big
data researchers often use Twitter to gauge public opinion on hot issues or
learn more about consumer preferences; however, these data are
problematic because Twitter users are not comparable to the U.S.
population as a whole (Tufekci 2014).
Using sources such as Twitter presents other unique sampling and
selection issues. A common avenue of big data analysis using Twitter is
“hashtag analysis”; that is, using Twitter’s linking system (tagging a post
with a “#” connects that post with a live feed of all users tweeting about that
particular topic) to gauge public opinion on timely, hot button issues.
Research by Tufekci (2014) highlights the inherent problem of selecting
cases based on the dependent variable. That is, users are only able to be
included in the sample if they have tagged their tweet appropriately. This
issue causes researchers to overlook cases where the user has not linked the
post but is still engaged in the larger conversation, and is necessarily subject
to self-selection bias, as the user has made a conscious choice to tag their
post and include their content in the larger discussion (Tufekci 2014;
Geddes 1990). Additionally, those hashtags that are used for analysis are
those that were successful (/%., generated a large base of users engaging in
Fall 2015
34
the conversation), and are therefore different than those hashtags that were
unsuccessful in generating conversation.
There are several ways researchers may account for selection bias in big
data analysis. Traditionally, selection biases are controlled for through
statistical modeling. Researchers should examine the overall demographics
of their data source and attempt to control for confounding variables.
Certain models and statistical methods are also better equipped to handle
such biases, such as a regression discontinuity design (Taylor 2014),
difference-in-difference models with a matching model as suggested by
Heckman (1979), and a non-linear Tobit model, used by both Heckman
(1979) and Berk (1983).
Additionally not all selection bias is problematic. In some cases, notably
market research, selection bias enables greater targeting of advertising. For
example, retailers use “just-in-time” coupon delivery to target specific
buyers. Entertainment services like Netflix and Amazon use similar
methods to make suggestions to viewers. Likewise, in research projects
where motivated participants are desired, the act of participating in a
hashtag conversation, itself may be noteworthy.
In the example of self-selection in hashtag analysis using Twitter,
researchers can better account for the confounding variable issues present
in a self-selected sample by going beyond exploratory, correlational
analyses. Twitter datasets should not be considered random or
representative; rather these data should be recognized as self-selected and
missing data treated as “missing not at random”, or missing due to
unobserved or unknown variables. In these analyses, it is also worthwhile
for researchers to examine the cultural and social contexts of these “trending
topics” and interpret their findings appropriately (Tufekci 2014; Meiman
and Freund 2012).
Additionally, big data research projects can be strengthened by pulling
dependent variables from external, validated sources. For example, a study
examining political attitudes using Twitter or Facebook content data as
independent or control variables could be strengthened by using voting
behavior or voting registration data from the U.S. Census as a dependent
variable. This strategy, in particular, is useful in guarding against “selecting
on the dependent variable”, as discussed by Tufekci (2014). Alternatively,
researchers may use outside, reliable sources, like the U.S. Census, to
Washington Academy of Sciences
35
benchmark their findings, acting as a sort of “gut check” for the findings. It
may also be worthwhile for big data researchers to explore mixed-methods
approaches to their studies. By complementing big data analyses with
surveys, interviews, and other data collection methods, researchers can
better understand the larger context of their data and provide more robust,
representative findings.
Big data research seems poised to revolutionize data analytics as we
know it. By amassing such large amounts of data, researchers can observe
correlations that may not manifest in smaller samples and can analyze large,
near real time streams of data from numerous sources. While detecting
previously missed correlations can spur new research questions and new
understandings of processes in fields like health and social science, it is not
a substitute for established theory, hypothesis development grounded in the
extant social science literature, and statistical modeling. These data, as
shown through this paper, may also be subject to selection biases, which
can skew the findings and implications of the research. By incorporating
more “small data” methods and techniques into research, such as mixed-
method studies, alternate non-linear models, and benchmarking, big data
analysts can strengthen their studies and findings and advance big data
research.
References
Anderson, Chris. 2008. "The End of Theory: The Data Deluge Makes the Scientific
Method Obsolete." Wired 16-07.
Berk, Richard A. 1983. “An Introduction to Sample Selection Bias in Sociological Data”.
American Sociological Review , 48(3), 386-398.
Bollier, David, and C. M. Firestone. 2010. The Promise and Peril of Big Data.
Washington, D.C.: Aspen Institute, Communications and Society Program.
De Mauro, Andrea, Marco Greco, and Michele Grimaldi. 2015. “What is Big Data? A
Consensual Definition and a Review of Key Research Topics”. In A IP Conference
Proceedings (Vol. 1644, pp. 97-104).
Duggan, Maeve, Nicole B. Ellison, Cliff Lampe, Amanda Lenhart, and Mary Madden.
2015. “Social Media Update 2014,” Pew Research Center. [Available at:
http://www.pewinternet.oru/20 15/01 /09/social-media-update-20 1 4/ (Accessed 10
September 2015)]
Fan, Jianqing, Fang Flan, and Han Fiu. 2014. “Challenges of Big Data Analysis”.
National Science Review, 1(2), 293-314.
Fall 2015
36
Geddes, Barbara. 1990. "How the Cases You Choose Affect the Answers You Get:
Selection Bias in Comparative Politics." Political Analysis 2: 131-150.
Harford, Tim. 2014. “Big data: A big mistake?” Financial Times, 11(5), 14-19.
Heckman, James J. 1979. “Sample Selection Bias as a Specification Error”.
Econometrica 47(1) 153-161.
Heckman, James, Hidehiko Ichimura, Jeffrey Smith, and Petra Todd. 1998.
“Characterizing Selection Bias Using Experimental Data”. Econometrica, 66(5),
1017-1098.
LaValle, Steve, Eric Lesser, Rebecca Shockley, Michael S. Hopkins, and Nina
Kruschwitz. 2013. “Big data, Analytics and the Path from Insights to Value”. MIT
Sloan Management Review, 2 1 .
Lazer, David, Ryan Kennedy, Gary King, and Alessandro Vespignani. 2014. “The
Parable of Google Flu: Traps in Big Data Analysis”. Science 343(6176), 1203-1205.
Lohr, Steve. 2012. “The Age of Big Data”. New York Times, Feb. 11 2012.
Meiman, Jon, and Jeff E. Freund. 2012. "Large Data Sets in Primary Care Research." The
Annals of Family Medicine 10(5) 473-474.
Pew Research Center. 2015. “The Smartphone Difference” [Available at:
http://www.pewintemet.org/20 1 5/04/0 l/us-smartphone-use-in-20 1 5/ (Accessed 10
September 20 1 5)]
Price, Megan, and Patrick Ball. 2014. “Big Data, Selection Bias, and the Statistical
Patterns of Mortality in Conflict”. SAIS Review of International Affairs, 34(1), 9-20.
Raghupathi, Wullianallur, and Viju Raghupathi. 2014. “Big Data Analytics in Healthcare:
Promise and Potential”. Health Information Science and Systems, 2(1)
[http://www.hissjoumal.eom/content/2/l/3 (accessed 10 September 2015)]
Sagiroglu, Serfef and Sinanc Duygu. 2013. “Big Data: A Review”. Proceedings for the
Institute of Electrical and Electronics Engineers 2013 Annual Conference. 42-47
Taylor, Eric. 2014. “Spending More of the School Day in Math Class: Evidence from a
Regression Discontinuity in Middle School”. Journal of Public Economics, 1 17,
162-181.
Tufekci, Zeynep. 2014. “Big Questions for Social Media Big Data: Representativeness,
Validity and Other Methodological Pitfalls”. In ICWSM ’ 14: Proceedings of the 8th
International AAAI Conference on Weblogs and Social Media. 504-514.
Yang, Qiang, and Xindong Wu. 2006. “10 Challenging Problems in Data Mining
Research”. International Journal of Information Technology & Decision Making,
5(04), 597-604.
Washington Academy of Sciences
37
Bios
Katie Seely-Gant is an analyst with the Analytics Team at the Energetics
Technology Center in Waldorf, Maryland (U.S.). Her work involves a wide
variety of projects in areas including science and technology workforce and
career pathways, diversity and inclusion in technical workforces,
international and domestic teams, and mentoring.
Lisa M. Frehill is Senior Analyst and Acting Director of the Analytics
Team at the Energetics Technology Center (Waldorf, Maryland, U.S.). She
is currently on detail as Organizational Evaluation and Assessment
Researcher at the National Science Foundation. Dr. Frehill is an
internationally recognized expert on human resources in science and
engineering, designing and executing program evaluations, strategic
workforce planning, and change management.
Fall 2015
Washington Academy of Sciences
39
Educating Data Scientists
in the Broader Implications of their Work
Heikki Topi and M. Lynne Markus
Bentley University, Waltham, MA
Abstract
The number of degree programs in analytics and data science is increasing
rapidly. Because of the strong industry demand for highly qualified analytics
professionals, the need for education will continue to grow. Current programs
provide strong coverage of the infrastructure and applications of analytics and
data science, but they are lacking in the coverage of their legal, ethical, and
societal implications. We argue that every analytics and data science program
should include a significant emphasis on the implications and potential
consequences of data science applications. Including these elements in the
programs will help analytics professionals understand better the complex and
nuanced relationships between their work and various stakeholders of the
context in which the work takes place. Data scientists and analysts who are
sensitive to data downsides as well as upsides enable organizations to avoid
harmful consequences of analytics applications but still achieve the benefits.
Introduction
Nothing is more crucial to the achieving the promises of big data than
a workforce of capable individuals prepared to tackle the opportunities and
challenges of analytics. Many commentators have mentioned projected
shortfalls in the number of people qualified to fill the data scientist role
(Craig et al. 2012, 2013; Manyika et al. 2011). A few analysts have also
pointed to the need for technical support specialists who manage data
preparation, storage, and related tasks (Woo 2013). Yet others have
discussed deficiencies in the ability of managers and subject matter experts
to sponsor, supervise, and take action on the results of analytic projects
(Court 2015; Davenport 2013).
The discussion has, however, paid less attention to what these
people need to know. Through our participation in NSF-sponsored
workshops on data science education1 and big data’s social, economic and
workforce implications,2 we have identified important knowledge gaps
related to the legal, ethical, and social implications of data science. In this
contribution to the Symposium, we lay out the current state of data science
Fall 2015
40
education and make the case for more attention to big data’s implications
and consequences in data science education.
Data Science Degree Programs
The demand for professionals with proficiency in data science and
analytic techniques has increased substantially during the last few years,
and many media accounts have predicted a significant shortage of capable
professionals in this area. A widely cited McKinsey report (Manyika et al.
2011) predicted a shortfall of nearly 200,000 knowledge professionals with
in-depth preparation in analytics. Industry demand for graduates with this
background has increased, and universities around the world have
responded by launching both bachelor’s and master’s programs.
North Carolina State University’s Institute for Advanced Analytics,
an educational pioneer, maintains a database of related U.S. master’s
programs. In September 2015, this list included 34 programs in Analytics,
19 in Data Science, and 54 in Business Analytics. Exact numbers of students
are difficult to estimate, but program heads frequently boast of their success
in attracting students from around the world. The largest programs admit
hundreds of students annually, and all appear to bring in at least dozens. In
total, data science and analytics programs already graduate thousands of
students per year in the U.S. alone. The focus on data scientist preparation
is not, however, only a U.S. phenomenon; new programs are also launching
in Europe and Asia. We confidently expect the number of analytics
graduates to continue to increase for some time.
Data Science Curricula
The disciplinary focus and orientation of data science and analytics
programs vary significantly, and so do their curricula. University
departments of statistics, mathematics, computer science, information
systems, management science/operations research, and information science
have set up programs, and so have schools and departments that focus on
particular sectors, such as health care, finance, and the hard sciences. Some
programs are interdisciplinary, crossing several departments or schools.
Other programs ( e.g ., most, but not all, business analytics programs) are
housed in single schools.
Washington Academy of Sciences
41
Some data science and analytics programs focus primarily on the
application of data science techniques to real world problems in science,
engineering, health care, financial, educational policy, and the like. These
programs tend to heavily emphasize data analytic techniques such as
traditional statistical analysis, data mining, text mining, time series analysis,
simulation, optimization, and machine learning, as well as domain
knowledge. In addition, because graduates are expected to work closely
with subject matter experts, these programs often include attention to
general professional competences such as critical thinking, oral and written
communication, collaboration and teamwork, and consulting skills for
eliciting project requirements. The applications of data science increasingly
involve continuous, real-time, algorithmic analysis of large quantities of
data in a way that enables automated organizational decision-making.
Therefore, some analytics programs (particularly in business schools) focus
on the role of analytics in the digital transformation of organizations.
Other data science programs focus more heavily on the activities
involved in providing the underlying support for data science and analytics
applications; these programs can be described as building skills in the
technical infrastructure of data science. Examples of the topics emphasized
in these programs are programming, algorithms and data structures, data
visualization approaches, data warehousing, data management for
structured and unstructured data, and so forth.
A Missing Emphasis
In addition to applications and infrastructure, a third key body of
knowledge relevant to data science and analytics concerns their legal,
ethical, and societal implications (see Table 1 that illustrates the three
bodies of Data Science knowledge based on Markus and Topi, 2015). One
can hardly mention the topic of big data without evoking privacy concerns,
and discussion of security issues often follows closely behind. Also relevant
are concerns about illegal discrimination, behavioral manipulation,
harassment, and inappropriate “social sorting” (or labeling people via
identity profiles) (Markus and Topi 2015). These legal, ethical, and societal
implications of big data are rarely given attention proportionate to their
importance in data science, analytics programs, or educational materials
(Provost and Fawcett 2013). Instead, such topics are typically covered in
educational programs of law, accounting information systems, social
Fall 2015
42
sciences, and public policy, where data scientists-in-training may not be
exposed to them.
Table 1: Three Bodies of Data Science Knowledge
(adapted from Markus and Topi, 2015)
Applications
• Use of data science tools and techniques to generate new insights in
domains such as marketing, health care, law, finance, science
• Use of data science tools and techniques to fully or partially
automate previously manual decision-making processes such as the
auctioning of advertising, mortgage or insurance underwriting,
medical diagnosis, e-discovery, securities trading, identification of
promising drug molecules
• Development of new “apps” and data-oriented business processes
and business models
Infrastructure
• Development of new tools and techniques for data handling (e.g.,
extracting, transferring and loading data, data storage, tagging and
curating data, cleaning and verifying data)
• Development of new software and hardware tools and techniques
for data analysis and interpretation (e.g., text mining, data
visualization, machine learning)
Implications
• Laws and regulations governing data protection, data security, and
data management requirements (e.g., document retention and
destruction)
• Design of organizational structures and governance mechanisms
that promote responsible (legal, ethical, and socially acceptable)
data collection and use practices
• Potential positive and negatives consequences of big data
applications, tools for anticipating them, and strategies and
techniques for minimizing negative side-effects.
Washington Academy of Sciences
43
Most data science and analytics degree programs appear to take~
either implicitly or explicitly — a strong pro-innovation stance, implying
that the consequences of big data are uniformly positive or, if negative,
easily remediable. We believe that this stance creates important gaps in the
preparation of the future data science and analytics professionals. In
particular, we believe that they are underprepared for the significant ethical
challenges they are likely to confront throughout their careers.
Why an Implications Focus Is Essential
Over the last century, thoughtful scientists, engineers, and technical
professionals have taken powerful stands on the uncertain or potentially
negative consequences of innovations such as fossil and atomic energy,
genetic engineering, and nanotechnology. Appropriate responses to such
issues are never easy to find and are always contested, but nothing is gained
by sweeping the issues under the rug. Failure to raise concerns and debate
the issues publicly generally only convinces the public that there is
something to hide and creates opposition that can block beneficial
innovations. This is as true of big data as it is of nuclear power. For instance,
the InBloom big data educational innovation was terminated after parents
voiced fears over possible secondary uses of their children’s data (Kharif
2014). When professionals are primed to understand and raise questions
about the possible downsides of a proposed data innovation, better
technology uses and outcomes are possible for all.
Among the reasons for preparing data scientists to understand the
broader implications of their work are the following:
• The systems that feed and flow from data analytics are often highly
complex, drawing data from numerous sources both external and
internal to an organization and involving interconnections among
independently developed systems. The sheer complexity of such
systems can give rise to unexpected outcomes and glitches.
Education is needed to anticipate, prevent, diagnose, and correct
such outcomes.
• Relying on intuition in the interpretation of analytics results can lead
to serious practical errors. Data scientists need to be well attuned to
the sources of error in data and algorithms and to human reactions
to labeling, subtle guidance, and constraints on their behavior.
Fall 2015
44
• Individual perspectives and value judgments regarding the
implications of analytics-based systems and the potential that
analytics infrastructures create vary quite significantly. Education
can help analytics professionals leam ways to put personal biases
aside and arrive at a more neutral analysis and a more successful
resolution of complex sociotechnical situations.
• Automated decision-making systems that operate without
continuous human involvement can amplify the negative
consequences of flawed data analyses. “Invisible technical workers”
(Ribes etal. 2013) make choices — both at the time of original design
and implementation of the systems and during system operation
(which might be fully integrated) — that potentially have far-
reaching consequences for both individuals and organizations.
These consequences are often opaque to organizational clients and
users. In some cases, even technical specialists may not understand
why their algorithms produce the results they do. Greater awareness
of this possibility is needed to produce and update algorithms that
work well, to devise effective ways for humans to intervene, and to
give affected people the opportunity for redress when errors are
made.
In short, realizing the potential benefits of big data without the possible
harms requires data scientists and analysts who are sensitive to data
downsides as well as data upsides.
Knowledge Areas and Pedagogical Approaches
We believe that data science and analytics programs need modules
and course(s) on the implications and potential consequences of analytics.
These courses must be specialized to the particular legal, ethical and societal
issues raised by big data. For example, simply adding a course on business
ethics (which may cover foreign corrupt practices, abusive labor practices,
and environmental damage) to a business analytics program does not
rigorously address issues like data protection law, personal information
privacy, online harassment, or glitches in online trading. Naturally, ethical
theories and general principles are shared across contexts, but the way these
principles are applied requires in-depth understanding of the dependencies
and connections discussed above.
Washington Academy of Sciences
45
It is beyond the scope of this paper to present a detailed proposal for
courses (or course modules) on the legal, ethical, and societal implications
of analytics and data science. However, we propose that the following
knowledge areas should be covered:
• Categories of major implications and potential consequences of
analytics-based systems (as laid out, for example, in the framework
proposed by Markus and Topi 2015)
• Methods for identifying, analyzing, and understanding complex
organizational situations from multiple perspectives in an unbiased
way, particularly from the perspective of implications and potential
consequences of technology-enabled systems
• Rich collections of relevant real-world examples that illustrate the
positive impacts of thorough implications analysis
• Material that allows students to understand themselves as ethical
decision-makers. It is particularly important that the students have a
good understanding of the sources of potential biases in data,
algorithms, and decision-making processes.
Pedagogically, the modules or courses required for developing these
competences need a good balance between the elements that build a strong
conceptual foundation and those that apply methods of active, participatory
learning. It is essential to allow students to internalize the theories and make
personal discoveries through case analysis, interviews and observations in
organizational settings, role play, games, simulations, and other similar
pedagogical approaches. The modules also need exercises that help the
students discover their own value positions on analytics issues. Many of
these materials do not currently exist, and developing them should be an
important priority for the data science and analytics community.
Conclusion
In this paper, we have argued that every analytics and data science
program should include a significant emphasis on the implications and
potential consequences of big data. This can be accomplished through a set
of components (or a single course) that provide students with the
opportunity to engage, theoretically and experientially, with the legal,
ethical and societal implications, and potential consequences of big data.
Without a systematic approach to developing these competencies, even
highly trained and technically competent experts may approach their work
Fall 2015
46
with perspectives that are too narrowly focused on the potential benefits of
innovation and too neglectful of potential harms. We firmly believe that
integrating attention to implications and consequences into every analytics
and data science program will lead to great value for individuals,
organizations, and society.
Acknowledgement
This material is based upon work supported by the National Science
Foundation under Grants No. 1348929 and 1545135.
Endnotes
1 Workshop on Data Science Education (National Science Foundation
award DUE: 1545135), Heikki Topi, Principal Investigator and Lillian
(Boots) Cassel, Co-Principal Investigator.
2 Big Data, Big Decisions — A Research Agenda Setting Workshop on the
Social, Economic, and Workforce Consequences of Big Data (National
Science Foundation award IIS: 1348929), M. Lynne Markus, Principal
Investigator.
Washington Academy of Sciences
47
References
Court, David. 2015. “Getting Big Impact from Big Data.” McKinsey Quarterly
(January).
[http://www.mckinsey.com/insights/business_technology/getting_big_impact_
from_big_data?cid=other-eml-alt-mkq-mck-oth-1501 (accessed September 25
2015)]
Craig, Elizabeth, Charlene Hou, and Brian F. McCarthy. 2012. The Looming Global
Analytics Talent Mismatch in Insurance. Accenture.
[https://www.accenture.com/us-en/insight-looming-global-analylics-talent-
mismatch-insurance.aspx (accessed September 25 2015)]
Craig, Elizabeth, Charlene Hou, and Brian F. McCarthy. 2013. The Looming Global
Analytics Talent Mismatch in Banking. Accenture.
[https://www.accenture.com/us-en/insight- global-analytics-shortage-banking-
summary. aspx (accessed September 25 2015)]
Davenport, Thomas H. 2013. “Keep Up with Your Quants.” Harvard Business Review
(July-August): 120-123.
Kharif, Olga. 2014. “Privacy Fears Over Student Data Tracking Fead to InBloom's
Shutdown.” Bloomberg Business.
[http://www.bloomberg.eom/bw/articles/2014-05-01/inbloom-shuts-down-
amid-privacy-fears-over-student-data-tracking (accessed September 6 2015)]
Manyika, James, Michael Chui, Brad Brown, Jacques Bughin, Richard Dobbs, Charles
Roxburgh and Angela Hung Byers. 2011. Big Data: The Next Frontier for
Innovation, Competition, and Productivity. McKinsey Global Institute.
[http://www.mckinsey.com/
insights/business_technology/big_data_the_next_frontier_for_innovation
(accessed September 25 2015)]
Markus, M. Fynne and Heikki Topi. 2015. Big Data, Big Decisions for Science,
Society, and Business- — NSF Project Outcomes Report. Waltham, MA:
Bentley University.
Provost, Foster and Tom Fawcett. 2013. Data Science for Business: What You Need to
Know about Data Mining and Data-Analytic Thinking. Sebastopol, CA:
O'Reilly.
Ribes, David, Steven Jackson, Stuart Geiger, Matthew Burton, and Thomas Finholt.
2013. “Artifacts that Organize: Delegation in the Distributed Organization.”
Information and Organization, 23: 1-14.
Woo, Ben. 2013. “Combating the Big Data Skills Shortage.” Forbes, 18 Jan.
[http://www.forbes.eom/sites/bwoo/2013/01/18/combating-the-big-data-skills-
shortage/ (accessed September 25 2015)]
Fall 2015
48
Bio
Heikki Topi is Professor of Computer Information Systems at Bentley
University in Waltham, MA. He has contributed to international
computing curriculum development and evaluation efforts in various
leadership roles since early 2000s (including CC2005 Overview Report;
task force co-chair of IS 2010, the latest undergraduate IS curriculum
revision; task force co-chair of MSIS 2016, ongoing revision of the IS
master’s curriculum; and ongoing ACM Education Council initiative on
Data Science education). He is co-author of a leading data management
textbook Modern Database Management, now in its 12th edition. His co-
edited Volume 2 Information Systems and Technology of CRC/Chapman
& Hall’s Computing Handbook was published in May 2014. He has been
a member of ACM’s Education Board since Spring 2006 and represented
first AIS and then ACM on CSAB’s Board since 2005.
M. Lynne Markus is the John W. Poduska, Sr. Professor of Information
and Process Management at Bentley University, Visiting Professor at the
London School of Economics, and Research Affiliate at MIT Sloan
School’s Center for Information Systems Research. She was the Principal
Investigator of an NSF workshop award to develop a research agenda on
Big Data’s social, economic, and workforce consequences, and she
participated in the White House 90-day review of Big Data. Markus was
named a Fellow of the Association for Information Systems in 2004 and
received the AIS LEO Award for Exceptional Lifetime Achievement in
Information Systems in 2008.
Washington Academy of Sciences
49
Everything Old is New Again: The Big Data
Workforce
Lisa M. Frehill
Energetics Technology Center, Waldorf, MD
Abstract
“Big Data” is a relatively new term, often used imprecisely and often in contexts
that imply a pressing need for workers with a newly-blended unique skillset.
However, social scientists have worked with exceptionally large data sets for
quite some time, historically accessing remote space, writing code, analyzing
data, and then telling stories about human social behavior from these complex
sources. Therefore, more than a half century of accumulated social science
knowledge about extracting information from very large data sets to understand
human social behavior provides a model for the emergent data science
profession. In this article I present analyses of current and projected U.S.
workforce data using various definitions of skillsets for data scientists,
concluding with a discussion of the policy implications.
Introduction
“Big Data” is a relatively new term, often used imprecisely. Recent
U.S. science and technology policy panels, including the President’s
Council of Advisors on Science and Technology (PCAST 2015) and a U.S.
National Academies of Science study group, deploy the term “big data” in
a way that suggests that in the past, the issues of privacy, worker skills, and
access were not encountered and are, therefore, new and in need of
attention. However, social scientists have worked with exceptionally large
data sets for quite some time, including implementing some of the much-
touted benefits of big data such as merging multiple datasets from disparate
sources into larger files, hierarchical file structures, and integrating
quantitative and qualitative sources to derive insights into human social
behavior. Hence, the methodological practices, ethical guidelines, data
management, and statistical methods that have been honed in the social
sciences' provide important workforce development lessons for the “new”
big data.
Additionally, depending on how one specifies the workforce skills
requirements for big data, the size of the existing pool of talent varies, as
does the prognosis for the labor market fortunes for data scientists as the
Fall 2015
50
21st century’s “sexiest job” (Davenport & Patil 2012, RJMetrics 2015). In
this article I start by differentiating the new big data as a form of what
Groves (2011) describes as “organic” data, in contrast to the traditional
large “designed” datasets that have been / continue to be collected, analyzed
and reported on by social scientists. Next I will present analyses of current
and projected U.S. workforce data using various definitions of skillsets for
data scientists. Finally, I close by discussing the policy implications about
the big data workforce.
What is Big Data and How Does it Differ from Previous Types of
Large Datasets?
The term “Big Data” typically refers to what Groves of the U.S.
Census Bureau has referred to as “organic” data (2011), in contrast to
“designed data.” As these terms imply, designed data are the traditional raw
material deployed by social scientists to answer research questions with
carefully designed studies using tested and accepted methodologies to
advance knowledge of the social world. Organic data are observational data
generated by the day-to-day behaviors of people. Social scientists also
gather organic data, but, again, such data collections are designed as
opposed to the data mining approach associated with the new big data.
Social scientists from many fields have worked with very large
datasets. The computing power available in personal computers now far
exceeds the capabilities of these machines just two decades ago. Those who
worked with very large files such as those from the U.S. Census, the Panel
Study of Income Dynamics, or the Department of Education’s High School
and Beyond longitudinal study program (to name just a few), often accessed
tape or cartridge-stored files using mainframe operating systems such as the
IBM VAX. Researchers’ individual accounts at colleges and universities
were typically insufficient in size to permit storage and analysis,
necessitating that social scientists who were “quants” learn how to access
and assign virtual space on which to park files and then perform statistical
analyses using one of a number of statistical packages like BMDP, SPSS,
or SAS (among others). Additionally, social scientists needed to learn the
command syntax for these specialized packages, therefore, rudimentary
programming skills were also a critical element of quantitative social
scientists’ training.
Washington Academy of Sciences
51
In the 1970s- 1990s, quantitative social scientists’ graduate
programs typically included advanced courses in research design, including
statistical analysis, but often did not emphasize visualizations, which were
not as commonly deployed at that time in peer-reviewed research literature.
Technical skills such as programming and command syntax were easily
learned via workshops offered by campus computer centers, basic classes
offered by computer science or management information science programs,
or programming manuals. The academic program focused on the substance
and design issues that provided a foundation for the analyses that social
scientists would perform on the large datasets. Quantitative social
scientists — in accordance with the deductive scientific method — started
with ideas and then located the appropriate data, which could be used to
answer research questions.
With the substantial increases in desktop (and laptop) computing
power, many social scientists now have the luxury of being able to store and
analyze many of these same datasets on a local machine rather than
mounting tapes and using remote computers. Additionally, many of the
popular statistics packages developed windows-based products, which
enable many quantitative social scientists today to more easily manipulate
and analyze data without learning complicated command syntaxes.
Turning to a consideration of the implications of big data with
respect to observational data, three features now make
organic/observational data “big”: velocity, variety, and volume, the three
V’s (McAfee and Brynjolfsson 2012; De Mauro et al. 2015). Table 1
compares social science designed observational data and organic “big” data
on the three Vs. In a nutshell, while the variety of designed and organic data
are vast, the large organic designed data used by social scientists are orders
of magnitude smaller than organic big data. Information derived from
designed data requires more time to emerge than the pace with which the
insights from organic data are demanded in business settings for data-driven
decision making.
Fall 2015
52
Table 1. Observational Data - Designed and Organic (i.e., Big)
Many of the techniques, tools, and protocols developed by social
science research communities to manage and share large designed
datasets — including attention to the ethical issues associated with collecting
these data — hold important implications for the big data workforce. Since
1962 the Interuniversity Consortium for Political and Social Research
(ICPSR) at the University of Michigan Institute for Social Research, has
provided a repository for large social science datasets. ICPSR curates these
data, provides support to social scientists via training sessions in
quantitative methods, and has long-established and continuously evolving
standards for the storage and use of data with attention to the issues of
Washington Academy of Sciences
53
confidentiality and privacy. A 2014 PCAST report on privacy issues and
big data recognized the potential value that could be added by the social
sciences in its third recommendation:
With coordination and encouragement from OSTP, the
NITRD agencies should strengthen U.S. research in privacy-
related technologies and in the relevant areas of social
science that inform the successful application of those
technologies. (PCAST 2014: xiii)
The fourth recommendation indicates a need to incorporate
education about privacy issues into the education and training of
professionals who work with big data. Social science research methods
classes — typically required courses at both the undergraduate and graduate
levels — cover these issues. Additionally, online training platforms2
designed to educate and certify social, life, and medical scientists in issues
associated with human subjects in research represent another mechanism by
which those in disciplines that engage with big data-computer scientists,
marketing researchers, and machine language programmers-could be
educated in the complex issues associated with protection of individuals’
privacy.
Workforce Considerations - What are the Skills Needed for “Data
Scientists?”
Estimates of the needs for the big data workforce vary widely
because the skillset is a somewhat moving target. Starting in 2009,
Hammerbacher suggested that the new big data required a new occupation,
the data scientist, who, at Facebook, would be able to use a variety of
programming skills — Hadoop, R, and Python — to access space for the new
huge datasets and complete analyses. Similar emphasis on programming
skills and alignment with computing and information technology (IT)
disciplines was reflected in a 2010 PCAST report on Networking and
Information Technology (NIT) Research and Development (NITRD):
NIT is the dominant factor in America’s science and
technology employment, and that the gap between the
demand for NIT talent and the supply of that talent is and
will remain large. Increasing the number of graduates in NIT
Fall 2015
54
fields at all degree levels must be a national priority.
Fundamental changes in K-12 education are needed to
address this shortage. (PCAST 2010: 85)
Other sources in 2010 and 2011, though, indicate additional skills
beyond the technical computing skills cited by PCAST. A 2010 Economist
article reported on the new profession, suggesting that data scientists
“combine the skills of software programmer, statistician, and
storyteller/artist to extract nuggets of gold hidden under mountains of data.”
In 2011, McKinsey reported that the skills needed to exploit big data were
so disparate, that three types of workers would be needed:
Our research identifies three key types of talent required to
capture value from big data: deep analytical talent — people
with technical skills in statistics and machine learning, for
example, who are capable of analyzing large volumes of data
to derive business insights; data-savvy managers and
analysts who have the skills to be effective consumers of big
data insights — i.e., capable of posing the right questions for
analysis, interpreting and challenging the results, and
making appropriate decisions; and supporting technology
personnel who develop, implement, and maintain the
hardware and software tools such as databases and analytic
programs needed to make use of big data. (Manyika et al.
2011: 103)
As described above, these same skills are akin to those of
quantitative social scientists who used programming skills to manipulate
data and perform statistical analyses to extract information from very large
datasets. Everything old is new again: data science is a new version of
quantitative social science, but without the research foundation in human
behavior and the ethical standards of the social sciences. It is important to
recognize, however, that much of social science work with large datasets is
basic research (i.e., fundamental knowledge creation), while the extraction
of information from big data is applied research (i.e., enabling data-driven
decision making in work settings).
Most recently, however, some observers of the emergent data
scientist profession are less optimistic about the future of this occupation.
Washington Academy of Sciences
55
Darrow (2015) concludes, “Enjoy your fat salaries while you can data
scientists, because the rising tide of new talent and — gasp — automation will
take their toll.” The same Fortune article quotes Alex Cosmas of Booz Allen
Hamilton (BAH), who indicates that BAH hires analysts and then trains
them in the technical skills of data science: “We look for raw
inquisitiveness, the intellectual curiosity which will repay you tenfold.” As
in the past, the pool from which data scientists will be drawn is broader than
the pool of those trained in computing or information technology.
What is the Size of and Trends in the U.S. Data Science Workforce?
There is no question about the proliferation of new occupations
associated with computing and the burgeoning size of the information
technology (IT) workforce. The ubiquity of computing technology has
created the need for a host of workers in occupations that did not exist two
decades ago. The rapidity with which demand for workers with IT skills as
well as the variety of such skills have resulted in a number of mechanisms
by which workers obtain these skills and, as well, how employers obtain the
skilled workers they need. A recent report by RJMetrics (2015) used data
from Linkedln to estimate the number of data scientists (worldwide) to be
1 1,400, many of whom held advanced degrees.
Figure 1 shows U.S. Bureau of Labor Statistics (BLS) projections
for growth in a number of science and engineering occupations for the 20 1 2-
2022 period along with the actual growth in these occupations in the
previous ten-year period (2004-2014). Between 2004 and 2014, the number
of jobs in the U.S. economy grew by 5.1 percent, with substantially more
rapid growth in computing and mathematical sciences occupations,
including those associated with software development, which grew by 37
percent and 27 percent, respectively. Architecture and engineering
occupations barely grew, with a substantial contraction in hardware
engineering. Projections of growth for the 2012-2022 decade match this
pattern across occupations, with the most substantial growth projected for
computing and mathematical sciences, especially software, both outpacing
the 10.8 percent projected growth for the overall number of jobs.
Fall 2015
56
U.S. Labor Force Growth - Actual and Projected
□ Actual Growth, 2004-2014 ■ Projected Growth, 2012-2022
40%
30%
20%
-C
| 10%
C?
0%
-10%
-20%
Note: Software includes computer programming and software engineering, a subset of
"Comp./Math. Sci.; Hardware includes Computer Hardware Engineers, a subset of "Arch. & Eng."
Figure 1. Historical and Projected Demand for IT Workers
Source: Analysis of data from the Bureau of Labor Statistics, 2014. “Table 1.2 Employment by
detailed occupation, 2012 and projected 2022 (Numbers in thousands)” and Current Population
Survey AAT-series Table 1 1 for 2004-2014.
The plots in Figure 2 provides another way to understand the past
decade of change in these technical occupations in comparative perspective.
For both median weekly earnings of full time workers and the overall
number of workers in each occupational category, a change ratio was
computed as follows:
Change ratio
(Epcc, 2014 Epcc, 2004)
Eocc, 2004
(Etotal, 2014 ~ Etotal, 2004)
Etotal, 2004
The change ratio was computed for four categories of occupations
(denoted as occ in the subscripts in the equation): computer and
mathematical sciences, architects and engineers, software, and hardware.
The x-axis shows the employment change ratio, while the y-axis plots the
change ratio of median weekly earnings for these same four occupations. A
change ratio of unity would indicate that the change in the specific
Washington Academy of Sciences
57
occupation is on par with that in the rest of the economy, while ratios under
one show slower change and ratios greater than one indicate faster growth.
As shown in the Figure 2, only growth in software occupations outpaced
the U.S. in both numbers and median weekly earnings, while computer and
mathematical sciences outpaced the U.S. in earnings but not numbers. Both
architects and engineers, including computer hardware engineers, grew
slower than the U.S. economy over the same decade in terms of both
earnings and number of workers.
Figure 2. Earnings and Employment Change Ratios, 2004-2014
C/MS: Computer and mathematical sciences
Software: Software developers, applications, and systems software
Hardware: Computer hardware engineering
Arch & Eng: Architects and engineers
Source: Analysis of data from the Bureau of Labor Statistics, Current Population Survey AAT-
series, Table 1 1 2004 and 2014 (employment) and Table 39 2004 and 2014 (Median weekly
earnings for full time wage and salary workers).
Fall 2015
58
Figures 1 and 2 show data only for the IT occupations associated
with big data. It should be noted, though, that academic discipline silos
complicate definition of the big data workforce. On the one hand, as
described above, quantitative social scientists who work with very large
data sets to generate new, basic science knowledge of human social
behavior do not use the term “big data” to describe the work that they do.
Computer scientists have embraced the term “big data.” Online forums
emphasize the primacy of programming, which reinforces a professional
boundary on the skills associated with accessing and analyzing these
organic data. The popular pej oration of social science — clearly
demonstrated by U.S. Congress members’ frequent attacks on social science
projects funded by the National Science Foundation, for example —
reinforces this barrier. The emphasis on technical programming skills and
algorithm development has been suggested as a replacement for the theory
development process with respect to social data, with one observer claiming
that “the data deluge makes the scientific method obsolete.” (Anderson
2008).
Conclusion
Big data has abundant applications for business, health, and finance.
The ability to rapidly analyze exceptionally large data sets from multiple
sources to provide information to enable actions in real-time offers promise
in a range of areas. For example, big data may enable more precise dosing
of medications and has been used to develop sensor technology to determine
when a football player needs to be side-lined because of his/her heightened
risk of concussion. Consumers may experience more efficient service and
process efficiencies may yield lower prices for consumers and higher profits
for businesses.
The emergence and evolution of the data science occupation bears
on-going scrutiny. In just a few years, employers have seen the value
associated with a cadre of workers who have both technical skills as well as
the ability to tell a story with data. However, as noted by BAH’s Cosmas,
locating inquisitive analysts and then training them up in the technical skills
may be the likely direction that will be taken with this workforce. In this
case, the potential recruitment pool is far wider than graduates of computer
science programs and, indeed, computer science programs will need to
provide students with experiences that encourage inquisitiveness about
Washington Academy of Sciences
59
human social behavior and with robust training about privacy and
confidentiality on par with that in social science methods.
There are many potential benefits that may be derived from cross-
pollination between computing, on the technical side, and social sciences,
on the substantive side, to deploy big data as a tool for human advancement
beyond capitalist accumulation. Both sets of fields, however, need to be
wary of professional boundary heightening, which introduces
inefficiencies. Time, energy, and effort are needed to develop data science
as a truly transdisciplinary field that can yield both an advancement of basic
science knowledge about human social behavior as well as applied science
information for data-driven decision making in real world contexts. There
is a place in such a transdisciplinary field for both designed and organic
data, the latter of which may be more effectively translated into information
when there is thoughtful consideration of research questions, the literature
that informs those questions, and use of previously developed analytical
methodologies. Better translation of social science research into actionable
information may help diminish the challenges of its relevancy that have
plagued public funding of social science.
In the 1970s- 1990s, inquisitive social science practitioners
demonstrated that the secrets of accessing and analyzing very large datasets
were relatively easy to acquire; the current trends in big data analytics
suggest this to be similar with respect to data science now. While the
volume and velocity of basic research in the social sciences is smaller and
slower than in big data, the same variety of data sources and implications
for data quality -i.e., validity and reliability-are similar. So everything old
is new again; more than a half century of accumulated social science
knowledge about extracting information from very large data sets to
understand human social behavior provides a model for the emergent data
science profession.
Fall 2015
60
Endnote
1 Social sciences are taken in a broad sense to include fields categorized as
such by the National Science Foundation (e.g., anthropology, sociology,
psychology, political science, and economics) as well as fields that
deploy similar methods such as marketing and educational research.
2 These platforms include: the Collaborative Institutional Training
Initiative at https://www.citiprogram.org/: the National Institutes of
Health Protecting Human Research Participants at
https://phrp.nihtraining.com/users/login.php; and the FHI360 Research
Ethics Training Curriculum at
http://www.flri360.org/sites/all/libraries/webpages/fl~ii-retc2/index.html.
Washington Academy of Sciences
61
References
Anderson, Chris. (2008). “The End of Theory: The Data Deluge Makes the Scientific
Method Obsolete” Wired. Accessed online at
http://archive.wired.com/science/discoveries/inagazine/16-07/pb_theory/ (original
post: 06.23.08).
Darrow, Barb. (2015). “Data science is still white hot, but nothing lasts forever” Fortune
(online) accessed at http://fortune.com/2015/05/21/data-science-white-hot/ 30
August 2015.
Davenport, Thomas H. & D.J. Patil. 2012. Data Scientist: The Sexiest Job of the 21st
Century. Harvard Business Review. 90(10): 70-76.
De Mauro, Andrea; Marco Greco, and Michele Grimaldi. 2015. "What is big data? A
consensual definition and a review of key research topics" A/P Conference
Proceedings 1644: 97-104.
(http://scitation.aip.org/content/aip/proceeding/aipcp/10. 1063/1.4907823).
doi: 1 0. 1 063/1 .4907823 (https://dx.doi.org/ 10. 1 063%2F 1 .4907823).
Economist (2010). “Data, Data Everywhere” accessed online,
http://www.economist.com/node/15557443, originally posted 25 FE 2010.
Groves, Robert. (2011). “Designed Data” and “Organic Data” blog posted May 31, 201 1
accessed at http://directorsblog.blogs.census.gov/201 1/05/3 1/designed-data-and-
organic-data (accessed 28 August 2015).
Hammerbacher, Jeff. 2009. “Information Platforms and the Rise of the Data Scientist”
pp. 73-84 in Segaran, Toby, and Jeff Hammerbacher Beautiful Data : The Stories
behind Elegant Data Solutions. Sebastapol, CA: O’Reilly.
McAfee, Andrew, and Erik Brynjolfsson. 2012. “Big Data: The Management
Revolution” Harvard Business Review. (October 2012).
Manyika, James, Michael Chui, Brad Brown, Jacques Bughin, Richard Dobbs, Charles
Roxburgh, and Angela Hung Byers. 2011. “Big data: The next frontier for
innovation, competition, and productivity.” McKinsey Global Institute.
National Research Council. 2014. Training Students to Extract Value from Big Data:
Summary of a Workshop. Washington, DC: National Academies.
President’s Council of Advisors on Science and Technology (PCAST). 2015. “Big Data
and Dala-Intensive Computing.” Accessed online at
www.whitehouse.gov/ostp/pcast. (August 2015).
. 2014. “Big Data and Privacy: A Technological Perspective.” Accessed online at
www.whitehouse.gov/ostp/pcast. (May 2014)
. 2010. “Designing a Digital Future: Federally Funded Research and Development
in Networking and Information Technology” accessed online at
www.whitehouse.gov/ostp/pcast. (December 2010).
Fall 2015
62
RJMetrics. 2015. “The State of Data Science” RJMetrics, Benchmark Report Series
(October 2015). [https://rjmetrics.com/resources/reports/the-state-of-data-science/
(accessed 8 October 2015)].
U.S. Bureau of Labor Statistics. 2014. “Current Population Survey, AAT-series, Table
11. Employed persons by detailed occupation, sex, race, and Hispanic or Latino
ethnicity” [http://www.bls.gov/cps/tables.htm (accessed 8 October 2015)].
. 2004. “Current Population Survey, AAT-series, Table 1 1. Employed persons by
detailed occupation, sex, race, and Hispanic or Latino ethnicity”
[http://www.bls.gov/cps/tables.htm (accessed 10 February 2014)].
. 2014. “Current Population Survey, AAT-series Table 39. Median weekly earnings
of full-time wage and salary workers by detailed occupation and sex.”
[http://www.bls.gov/cps/tables.htm (10 February 2014)].
. 2004. “Current Population Survey, AAT-series Table 39. Median weekly earnings
of full-time wage and salary workers by detailed occupation and sex.”
[http://www.bls.gov/cps/tables.htm (10 February 2014)].
. 2014. “Table 1.2 Employment by detailed occupation, 2012 and projected 2022.”
[http://www.bls.gov/cps/tables.htm (10 February 2014)].
Bio
Lisa M. Frehill is Senior Analyst and Acting Director of the Analytics
Team at the Energetics Technology Center (Waldorf, Maryland, U.S.). She
is currently on detail as Organizational Evaluation and Assessment
Researcher at the National Science Foundation. Dr. Frehill is an
internationally recognized expert on human resources in science and
engineering, designing and executing program evaluations, strategic
workforce planning, and change management.
Washington Academy of Sciences
63
Social Media Analysis for Higher Education
Anamaria Berea, William Rand, Kevin Wittmer
University of Maryland
Gerard Wall
vibeffect
Abstract
The educational system involves a complex set of actors, including learners,
parents, teachers, and administrators. However, we now have more data than ever
to analyze this system, which could result in a quick understanding and evaluation
of public policies in this complex policy area. This paper explores a new area of
data about the educational experience, namely social media data. This paper
outlines an exploratory analysis of the Twitter discussions regarding higher
education in the USA. Based on a collection of more than 1.5 million tweets over
a period of 4 months, we identify a few key issues in the current higher education
discourse on social media. We also identify the effect of the expressed feelings of
the social media users when it comes to college applications, decisions and
completion. We conclude that policies in higher education can be better tailored
if they are informed by social media discussions.
Introduction
The increasing amount of data, the decreasing cost of computational
power, and the improving state of analytics has revolutionized fields from
stock trading to social analytics, but somehow higher education has not
received as much attention. The technology that has transformed many for-
profit businesses and governments can be applied at various colleges and
universities.
One obvious place that analytics could be useful is in the classroom,
but currently instructors at many universities are using outdated and
inefficient methods to grade assignments and compile these scores into self-
generated databases. In fact, Darnell West argues that “many of the typical
pedagogies provide little immediate feedback to students, require teachers
to spend hours grading routine assignments, are not very proactive about
showing students how to improve comprehension, and fail to take
advantage of digital resources that can improve the learning process” (West
2012). Data mining and analytics provide the capabilities necessary to
circumvent the traditionally cumbersome grading processes and glean
Fall 2015
64
insights from student data about performance, learning approaches, and
other metrics. For example, Leah Macfadyen and Shane Dawson developed
an “early warning system” which correctly identified 81% of students who
failed an online course by creating a regression model that analyzed such
variables as total number of discussion messages posted and total number
of assignments completed (Macfadyen and Dawson 2010).
Big data analytics within education could also be used to monitor
student progression through various course sequences for specific majors,
online courses that change activities by measuring everything from
individual clicks to aggregate performance and algorithms that suggest
courses a student should take by analyzing her past grades in similar courses
(Bienkowski etal. 2012). While traditional in-person classrooms may allow
for the collection of big data for these applications, Anthony Picciano notes
“to move into the more extensive and especially time-sensitive learning
analytics applications, it is important that instructional transactions are
collected as they occur” (Picciano 2012). This rapid collection of data is
most likely to be facilitated by course management/leaming management
system architectures and online and blended learning course structures
(Worsley 2012).
There is little work that has looked at how to use analytics methods
outside the classroom to improve the overall educational ecosystem, as well
as educational policy. However, insights produced by the previously
described learning analytics systems can also be used to inform policy
decisions. According to van Bameveld, Arnold, and Campbell (2012),
“Like business, higher education is adopting practices to ensure
organizational success at all levels by addressing questions about retention,
admissions, fund raising, and operational efficiency”. Michael Horn and
Katherine Mackey (2011) suggest that education analytics can be used to
shift the focus from inputs to outputs when measuring academic
institutional success. Instead of using seat-time, faculty-student ratios, and
dollars spent as a measure of success, analytics software can provide
information on more appropriate metrics such as student performance and
retention rates. The biggest obstacles to establishing more such systems are
building data sharing networks where these myriad metrics can be
aggregated, holistically analyzed, and shared among different institutions
(West 2012). A recent paper proposes a model and algorithm that would
Washington Academy of Sciences
65
help prospective students make better informed decisions about the best fit
and best college eco-system based on their unique personalities and
behaviors (Berea et al. 2015).
Text mining, social media, or sentiment analysis on the college
decision process has generally not been discussed in education analytics
literature and therefore presents an interesting opportunity to further
advance research in this area. A recent survey by Piper Jaffray found that
teens are abandoning Facebook in favor of Instagram; 76% of teens are on
Instagram and they are using it to gain an unfiltered look at colleges
(Stampler 2015).
Data Analysis
We collected data for this education analytics project for a period of
4 months, between March 4th and July 1st, 2015. For this collection we used
TwEater, an original and proprietary collection tool developed at the
University of Maryland (TwEater 2015). Originally, the collection was
based on 57 keywords and hashtags, such as: “igotin”, “college”, “campus”,
“acceptanceletter”, and many more, and the original data set comprised
more than 10 million tweets. Since most of these keywords were not
necessarily related to the idea of higher education and college admissions
and applications, we selected a list of 25 hashtags pertaining exclusively to
college, high school and higher education. Out of these, only 20 rendered
more than a tweet, with a minimum of one tweet for the hashtag
#choosingacollege and a maximum of 1,153,618 tweets for the hashtag
#college followed by 282,139 tweets for the hashtag #campus (see Table 1).
On the basis of this collection, we assembled a data set of 1,523,817
tweets where most of them (73%) refer to the general idea of “college”.
Many of these tweets are quite general, but some of them focus on specific
issues, such as: making college applications friendlier for the LGBT
community, businesses supporting campuses, parent-student conflicts in
college decision making, and hard college choices between various schools.
Text Mining
Based on this collection, we built a dictionary of about 470,000
unique words that are specific to the discourse about higher education in the
Fall 2015
66
USA. This is a dictionary roughly half the size of the English language (the
Oxford English Dictionary has over 600,000 words alone), with the caveat
that some of the words in our dictionary are informal or abbreviations or
pronouns that may not be currently recognized as being part of the formal
English.
The most frequent words in the education discourse are “campus”
and “college”, but if we leave these obvious terms aside, words such as
“highschool”, “acceptance”, “life”, and “met” are highlighted as the most
frequent ones that are not directly related to colleges. This gives us an
indication that students do talk about college acceptance, life, and college
related meetings on Twitter.
We also analyzed each of the 20 keywords separately and created a
histogram of word frequency for each of the 20 keywords. After “college”,
“campus”, “higher education” and “highschool”, the largest corpuses
(indicated by the number of tweets) belong to hashtags such as
#collegeopportunity, #collegetour and #collegebound. The second most
frequent word in most corpuses is “student”. Some interesting words, which
are sparse (low frequency) but appear more than once and are associated
with the most frequent terms mentioned above, are terms such as: “success”,
“community”, “hard”, “chip” and “app”. There is a very large gap between
the most frequent words and the second most frequent words (showing the
long tail distribution of the words) (see Table 1).
Twitter only allows for a fixed number of characters per tweet,
therefore we also checked how many unique words are being used in a tweet
in our data: #collegedecision, #collegechoice and #backtocollege have the
most “rich” tweets (an average of ~7 words per tweet), while
#collegeopportunity has the least number of unique words per tweet (an
average of 0.2), probably due to a different type of content used in the tweet
(z.e., hyperlink or video) (see Table 1).
Washington Academy of Sciences
67
Table 1. The summary statistics for higher education Twitter data
Sentiment Analysis
We matched the words in each of the 20 dictionaries with the
AFINN standard sentiment dictionary (Nielsen 2011) and calculated the
sentiment scores of the tweets in our data (see Figure 1). The AFINN
dictionary uses a scale from -5 to +5 to rate the effect of approximately 2000
words. We calculated both the absolute and the weighted scores for each
keyword. The absolute scores show that the first largest corpuses
(“campus”, “college” and “highschool”) are also strongly negative, while
all the rest are positive (with the exception of #backtocollege, where the
absolute score is only -1, close to neutral, and #collegedecision, which is 0).
However, raw sentiment scores do not take into account the volume of the
tweets for each keyword. We therefore examine weighted sentiment scores
- on corpus size and on tweet - since the distributions of the corpus sizes
and number of tweets are quite skewed. The weighted sentiment scores
show that #highschool is the most negative talk on Twitter, while
#collegematch and #collegeopportunity are the most positive ones.
Fall 2015
68
Absolute and Weighted Sentiment Values for Each Keyword
0)
(/)
Figure 1. Sentiment values for each keyword.
2.3. ZipPs and power law distributions
Zipf s law is a well-known statistical regularity observed in natural
language (Zipf 1949) that states that the frequency of any word is inversely
proportional with its’ rank in the frequency table. We tested whether the
Zipf law holds for each of the 20 corpuses and found that #highschool,
#highereducation, and #collegetalk have distributions similar to the Zipf
distribution (power of ~ -1), while #backtocollege, #rightcollege, and
#collegecompletion show the farthest departures from the Zipf distribution
(with power of ~ -0.3) (see Figure 2).
Power law distributions in college Twitter talk
Log Rank
Figure 2. Power law and Zipf distributions of words.
Washington Academy of Sciences
69
One way to interpret this result (backed by the size of the corpuses,
as well) is that there is more actual discussion involved in general topics
about high school and college education as opposed to topics about college
completion or matching, where the Twitter activity is more likely to inform
with links and other types of information as opposed to offering opinions
and personal insights and affect. There is no explanation today for why
Zipf s law is characteristic to human language, but some prior research
suggests that this distribution is more characteristic to natural language and
the human memory of language (Cohen et al. 1997; Piantadosi 2014).
Therefore tweets that contain other type of content than words are less likely
to exhibit this pattern.
Retweets
Re tweets in any Twitter data are one way to measure the degree of
popularity of certain tweets. In our data the retweets to tweets ratio is quite
high. The two keywords with the highest retweet to tweet ratio,
“rightcollege” and “collegeopportunity”, had retweeting activity for almost
each and every tweet — 0.933 and 0.903 respectively - but this is due to the
majority of the tweets with “collegeopportunity” that are initiated by the
users of @WhiteHouse and @BarackObama, which are popular and
frequently retweeted.
Disregarding these outliers, the two keywords with the highest
retweet to tweet ratio are “highered” and “collegebound” at 0.484 and
0.468, respectively - almost half of the tweets being retweeted. The high
retweet to tweet ratio of “collegebound” provides an interesting insight in
the context of this project. It indicates that many high school seniors revert
to Twitter to broadcast their accomplishments to friends, who share the
congratulatory experience. This conclusion is supported by a reading of the
tweets. Many of the keywords with high absolute numbers of tweets also
have moderately high retweet to tweet ratios, namely “highschool,”
“campus,” and “college” at 0.443, 0.399, and 0.356 respectively.
Conclusion
Our analysis is constrained to only about 4 months of collection and
a short list of keywords. But even so, our findings show the following: there
is generally a negative sentiment regarding colleges, campuses, high school
Fall 2015
70
and higher education; there is a tension between students and parents with
respect to college decisions; campuses and colleges are being judged with
respect to their inclusions ( i.e ., LGBT); people are more interested in
offering their opinions on general subjects (i.e., “campus”) than on specific
ones (i.e., “college tours”, “back to school”).
Our current research, although exploratory, points towards a few
general conclusions when using social media or Big Data for education
research. First, the selection of keywords and hashtags is essential, as these
are going to determine the constraints for the data that are going to inform
any analysis. Second, while there is considerable discussion on Twitter with
respect to higher education, most of this discussion is negative. Third, social
media is a great resource of information for education policy, as it gives in
real time the opinions of the parents and prospective students when it comes
to college applications, college acceptance, or college campuses.
Acknowledgements
The authors wish to thank Mrs. Elbe Cox for support and partnership
in initiating and conducting this research. The work has been entirely
supported by vibeffect.
References
Berea, Anamaria, Maksim Tsvetovat, Nathan Daun-Barnett, Mathew Greenwald, and
Elena Cox. 2015. "A new multi-dimensional conceptualization of individual
achievement in college." Decision Analytics 2(1): 1-15.
Bienkowski, Marie, Mingyu Feng, and Barbara Means. 2012. “Enhancing Teaching and
Learning Through Educational Data Mining and Learning Analytics.” United States
Department of Education Briefs [https://tech.ed.gov/wp-
content/uploads/2014/03/edm-la-brief.pdf (accessed 6 September 2015)]
Cohen, Albert, Rosario, N. Mantegna, and Shlomo. Havlin. 1997. “Numerical analysis of
word frequencies in artificial and natural language texts”. Fractals, 5(01): 95-104.
Horn, Michael B. and Katherine Mackey. 2011. “Moving from Inputs to Outputs to
Outcomes.” Innosight Institute, [http://www.christenseninstitute.org/wp-
Washington Academy of Sciences
71
content/uploads/20 13/04/Moving-from-Inputs-to-Outpuls-to-Outcomes. pdf
(accessed 7 September 2015)]
Macfadyen, Leah P. and Shane Dawson. 2010. “Mining LMS data to develop an ‘early
warning system’ for educators: A proof of concept.” Computers & Education 54(2):
588-599.
Nielsen, Fenn. 2011. “a new ANEW: Evaluation of a word list for sentiment analysis in
microblogs” in Proceedings of the ESWC201 1 Workshop on 'Making Sense of
Microposts': Big things come in small packages 718 in {CEUR} Workshop
Proceedings: 93-98. [http://arxiv.Org/abs/l 103.2903 (accessed 7 September 2015)]
Oxford English Dictionary. 2015. [http://public.oed.com/about/, (accessed 7 September
2015)]
Piantadosi, Steven T. 2014. “Zipf s word frequency law in natural language: A critical
review and future directions”. Psychonomic bulletin & review, 21(5): 1112-1130.
Picciano, Anthony G. 2012. “The Evolution of Big Data and Learning Analytics in
American Higher Education.” Journal of Asynchronous Learning Networks 16(3): 9-
20.
Oxford English Dictionary. 2015. [http://public.oed.com/about/, (accessed 7 September
2015)]
Rand, William, David Damron, and Radu Machedon, 2015. Tw Eater.
[https://github.coin/CenterForComplexitvInBusiness/ (accessed 7 September 2015)]
Stampler, Laura. 2015. “How High School Students Use Instagram to Help Pick a
College”, Time Magazine, [http://time.com/3762067/college-acceptance-instagram-
high-school/ (accessed 7 September 2015)]
Van Bameveld, Angela, Kimberly E. Arnold, and John P. Campbell. 2012. “Analytics in
Higher Education: Establishing a Common Language.” Educause Learning
Initiative : 2-11.
West, Darrell M. 2012. “Big Data for Education: Data Mining, Data Analytics, and Web
Dashboards.” Governance Studies at Brookings : 1-10.
Worsley, Marcelo. 2012. “Multimodal learning analytics: enabling the future of learning
through multimodal data analysis and interfaces.” in Proceedings of the 14th ACM
international conference on Multimodal interaction (ICMI '12). ACM, New York,
NY, 353-356. DOI=http://dx.doi.org/l 0.1 145/2388676.2388755
Zipf, George K. 1949. Human Behavior and the Principle of Least Effort. Cambridge,
Massachusetts: Addison-Wesley.
Fall 2015
72
Bios
Anamaria Berea is a postdoctoral researcher in the Center for Complexity
in Business, Robert H. Smith School of Business, University of Maryland.
She is researching social phenomena using various computational methods.
William Rand is an assistant professor of Marketing and Computer Science
at the University of Maryland. William Rand serves as the director at the
Center for Complexity in Business. His work examines the use of
computational modeling techniques, like agent-based modeling, geographic
information systems, social network analysis, and machine learning to help
understand and analyze complex systems, such as the diffusion of
innovation, organizational learning, and economic markets.
Kevin Wittmer is an undergraduate researcher in the Robert H. Smith
School of Business, University of Maryland. He is researching various
aspects of qualitative and quantitative methods for data analysis.
Gerard Wall is the Solutions Architect at vibeffect, pioneering the
investigation of how the Higher Education decision and its impact on
families can become more transparent and relevant for the “consumer' as
family.
Washington Academy of Sciences
73
Privacy in a Networked World:
New Challenges and Opportunities for Privacy
Research
Heng Xu and Haiyan Jia
The Pennsylvania State University
Abstract
In this article, we describe the new threats to information privacy that appear as
the result of the emerging Big Data practices and methodologies in today’s
networked world. In particular, the collection and analysis of large-scale data
from social networking sites challenge the traditional conceptualization of
privacy. In response, a new conceptual framework is proposed to encompass
three key dimensions of privacy in the Big Data context: information
identifiability, information ephemerality, and information linkability.
Introduction
The “privacy as a right” perspective, first introduced by Warren and
Brandeis (1890), has since influenced numerous opinions and court cases
on privacy and law enforcement (searches and seizures), privacy and self
(abortions and embryos), privacy and the press (private facts exposure and
celebrity privacy), privacy in the workplace (psychological testing and
lifestyle monitoring), etc. (Alderman and Kennedy, 1997). However, these
issues were just a subset of privacy issues which Warren and Brandeis were
concerned about when they wrote the “right to privacy.” Their main
concern was with the advent of technological developments (instant
photography and audio recordings in the late nineteenth century) that were
increasingly revealing personal information without individuals’
awareness.
Such privacy concerns still exist and remain highly relevant after 125
years. In today’s Big Data Era, many data collectors, data brokers,
aggregation services and various companies collect and use personal data
without individuals’ awareness, which leads to a dark data ecosystem. It has
been estimated that there are 4000 separate companies involved in the dark
data market and many dark data brokers make the data available to any
buyer willing to pay (Levine 2013).
Fall 2015
74
The emerging field of big data analytics is distinguished from
traditional data analytics by its three key characteristics (McAfee et al.
2012): (1) Volume - Big Data analysis works with petabytes of data in a
single dataset; (2) Velocity- Real-time or nearly real-time information is
aggregated and analyzed for agile decision-making; and (3) Variety - Big
Data takes all forms of information ranging from sensor readings and GPS
signals to messages, updates, and images posted on Social Networking Sites
(SNSs). Collecting large amounts of data, especially personal and social
data, brings both opportunities and challenges. While many practitioners
believe that the rise of Big Data has potential for creating better tools and
services, scholars ( e.g ., boyd & Crawford 2012) have already warned about
how poor execution may lead to negative social and economic
consequences such as intrusion to personal privacy, suppression to speech,
and misleading predictions, among many others.
Interaction between technological innovations and social ecology
usually has consequences far beyond the immediate purposes of the
technical devices and practices (Kranzberg 1986). One of the major threats
that Big Data analytics posits is privacy, as it seeks to identify at the
expense of individual and collective identity (Richards & King 2013).
Viewing Big Data as a public good, Acquisti (2014) discusses its critical
importance for public decision-making, and how it can reduce
inefficiencies and increase welfare when used properly. However, Acquisti
(2014) also questions who should bear the economic cost of Big Data
practices that use personal information: data subjects (whose data are
aggregated and analyzed), data holders (who collect and handle consumer
data), or both? To face the increasing costs associated with data storage and
analysis, data aggregators, and data holders typically assume that they have
rights to the data and exploit user data for profit, overriding the interests of
individuals in their privacy and leaving them few mitigating measures
(Wigan & Clarke 2013). Big Data practices as such pose significant threats
to individual privacy.
This paper aims at discussing the following challenges to information
privacy with the emergence of Big Data: (1) What are the unique threats of
Big Data practices to information privacy? (2) How do these unique threats
challenge the conceptualization of privacy? (3) How should we address
privacy challenges in today’s networked world?
Washington Academy of Sciences
75
Redefining Privacy in a Big Data Context
Information Identifiability
In today’s networked world, privacy is often shaped and enabled by
various features of technologies. For instance, websites often require users
to disclose certain types of information in order to obtain the services, and
provide certain mechanisms and tools for users to manage their privacy
preferences. Personal information, such as name, location, personal
interests, and even information of one’s social networks, can be revealed
voluntarily by the users to socialize and establish social connections.
However, disclosure of private information can have significant
consequences, and thus trigger users’ privacy concerns and shape their
privacy management behaviors (Fogel & Nehmad 2009). For instance,
popular SNSs such as Facebook and Twitter require various levels of
information disclosure and information accuracy by design. Facebook
requires users to provide real names and work/education email addresses to
be added to an affiliation or network. Twitter, on the other hand, does not
necessarily require real names, but it sets users’ profiles as public by
default, potentially exposing a large amount of personal information to the
wide audience and other third parties. Further, many social networking and
mobile applications monitor, record and even publish users’ location
information, which is susceptible to unauthorized disclosure.
In the context of Big Data, we argue that one fundamental dimension
of information privacy is information identifiability , which is defined as the
amount and the accuracy of personally identifiable information being
revealed. Unique to Big Data practices, individuals’ identities can be easily
identified or re-identified. For instance, Narayanan & Shmatikov (2009)
have demonstrated how to efficiently de-anonymize a large number of
Twitter and Flickr users by simply using data of username, location, and
“follow” or “contact” relationships. Data mining using vast amounts of
identifiable information generate hypotheses and discover general patterns
that could actually be stereotypical and misleading, possibly causing both
privacy loss and economic loss for data subjects, and posing privacy threats
that the existing privacy laws are far behind to define or protect (Brankovic
& Estivill-Castro 1999). What is more risky is that Big Data analytics can
now gather and extract implicit user data and across different social
networking platforms. Beyond user specified data such as usernames and
Fall 2015
76
locations, SNS services now can automatically generate user information
through mechanisms such as face recognition, geo-tagging, and multi-site
uploading, further increasing the amount and the accuracy of personal
information (Smith et al. 2012). These data are usually extracted from the
uploaded files or generated through metadata. Users are often unaware that
these data are stored and can be used for identification.
Thus, information identifiability is one key dimension of privacy, and
has extensive new meanings in the Big Data context. Big Data tools
significantly increase the potential to identify individual users through
social data and reveal more user information in increasing quantities and
accuracy. The lack of user awareness and regulatory mechanisms to control
such information revelation signifies its impact on information privacy.
Information Ephemerality
However, information identifiability does not fully capture the scope
of information privacy, especially in the Big Data era. Palen and Dourish
(2003) argue that privacy is not only about the identity boundaries defining
self versus others, but also the temporal boundaries between past, present
and future. Events of information disclosure are not isolated, but
sequentially connected. Therefore, information disclosed at a specific
instance becomes contextualized and interpreted in relation to other events
and situations, if the latter are available. In our daily life, information tends
to be ephemeral; the information that we share and exchange is constrained
to a certain physical location and a certain time period before it gets
forgotten. While we constantly observe the action of forgetting in our social
life (Mayer-Schonberger 2009) and in social norms and policy (Blanchette
& Johnson 2002), recent advances in information technologies have offered
inexpensive, large-volume digital data storage capacity, making the
persistence of information the odd commonplace (Ambrose 2012). The
extended information lifespan has significant privacy implications, as the
preservation of personal information amplifies and prolongs the effect of
any privacy loss. The persistence or the ephemerality of information has not
been a major privacy concern in the past few decades, but more recently, the
current and new consensus of privacy threat is formed around the fact that
information, once online, is there forever. This new realization has brought
attention to this new aspect of privacy — “the right to be forgotten”
Washington Academy of Sciences
77
(Ambrose 2012). We define information ephemerality as the duration of
private information being available, accessible and stable, an increasing
significant dimension of information privacy in close relation to big social
data.
While the traditional forms of data analytic tools may not be able to
handle large-scale longitudinal data, Big Data technologies, in particular,
can use the persisting records of social data, sometimes beyond a single
SNS platform, and change the availability and accessibility of information
from the “here and now” to the “everywhere and forever” (Grudin 2002).
The accumulated user data on Facebook alone have been used to reveal the
evolution of user interactions over three years (Viswanath et al. 2009), the
longitudinal changes in privacy and disclosure behaviors in six years
(Stutzman et al. 2012), as well as the year-long variation of national
happiness levels (Facebook 2010). However, changing the ephemeral
nature of information and making longitudinal analysis of such big social
data can be damaging. When modeling large datasets over time, many
time-sensitive factors may come into play to influence outcomes. Without
considering these factors and changes over the course of time, data will be
taken out of context, often lose meaning and value, and be interpreted in
misleading ways. For instance, boyd and Crawford (2012) point out that the
types of social networks derived from mining a longitudinal
dataset — “articulated networks” (networks resulted from people specifying
contacts through mechanisms such as friend lists or instant messenger lists)
and “behavioral networks” (networks derived from communication patterns
such as email exchanges and Facebook photo-tagging) — tend to be
inequivalent to true personal networks. “False discoveries” like this made
out of the large-scale social data not only breach personal privacy, but may
have severe real-world consequences affecting the products, bank loans,
and health insurance a person receives.
Information persistence is a unique “big social data” threat to users’
information privacy. Because the real world is one that is ephemeral rather
than permanent, individuals apply the same kind of expectations to their
online disclosure, expecting the information that they share online will not
be everlasting (Shein 2013). Big Data tools not only serve to document and
store longitudinal redords of private information, but also use and analyze
them for inferences, knowledge, and trends regarding users’ behavioral
Fall 2015
78
intentions and social implications, significantly challenging users’ privacy
expectations and violating their privacy rules.
Information Linkability
User data on SNSs concern not only information about the users
themselves, but also information about the users’ colleagues, friends, and
others they come into contact with. As SNSs facilitate connectedness across
boundaries and in dynamic ways, neither a one-time snapshot nor an
over-time trace of a single user’s profile can fully capture the complexity of
SNS data (boyd & Crawford 2012). Unique to the social data generated and
accumulated online, information privacy is dependent not on one single
user, but on a web of users to whom this individual is connected and on the
information that they disclose. Xu (2012) proposes the notion of privacy
2.0, describing this phenomenon that information disclosure is
co-constructed by users and their social connections, which demands the
responsibilities of privacy protection to be distributed through their social
networks. Following Xu (2012), we suggest that information linkability as
the third key dimension of information privacy in the Big Data context, and
define it as the degree to which information is relational and linked through
social connections.
As privacy scholars (Lampinen et al. 201 1; de Wolf et al. 2014) have
recently observed, the connected nature of SNS data and the interpersonal
nature of information sharing have made individualistic privacy protection
strategies inadequate. Even if a user adopts tight privacy settings, his or her
personal information could still be accessed or misused by their friends’
ignorance of privacy and security (Xu, 2012). As a result of such
information linkability, SNS data are often gathered and exploited without
the consent of the individuals to whom the data relate, and individuals who
volunteer such data only have moral responsibility for their actions (Wigan
& Clarke 2013).
Privacy risks in relation to information linkability become an
especially prominent problem with Big Data practices. Many analytic tools
are specifically designed for social network analysis to draw patterns and
insights from cliques, groups, and even large social networks (Davenport et
al. 2013). To address this emerging privacy issue, Troshynski et al. (2008)
argue that users, researchers, and practitioners of big social data consider
Washington Academy of Sciences
79
not only personal privacy implications but accountability, a broader concept
that encompasses the privacy considerations in a multi-directional
relationship - accountability for others’ personal privacy, accountability to
superiors, colleagues, and to the public. To achieve accountability,
considerations need to be given to the control of and power regarding the
access and use of linked and relational information and the differences
between accessibility and publicness (boyd & Crawford 2012).
The Emerging Field of Human-Data Interaction (HDI)
The technological advancement on machine learning and automated
content analysis continues to improve the strength of today’s big data
ecosystem. To address privacy concerns, many privacy scholars suggest
that individuals’ awareness of privacy should be enhanced by providing
information transparency about what data is collected, how it is used, and
whom it is shared with (Wang et al. 2013; Xu et al. 2012). In the privacy
literature, researchers have examined multiple ways of enhancing
transparency, such as providing explicit textual privacy statements (Pollach
2006), presenting privacy facts in the form of nutrition labels (Kelley et al.
2009), using warning icons to suggest suspicious data use (Lin et al. 2012),
and using justification messages to explain information disclosure
(Knijnenburg and Kobsa 2013). Transparency is also at heart of existing
and proposed regulatory schemes. For instance, the U.S. Consumer Privacy
Bill of Rights suggests that “companies should provide clear descriptions of
[...] why they need the data, how they will use it” (White House 2012).
While empowering individuals with privacy comprehensiveness is a
desirable approach to raise awareness, information transparency itself
cannot guarantee privacy. If implemented inappropriately, the strategies
can even backfire. For instance, practices to enhance information
transparency have been criticized for i) burdening users’ cognitive load by
having users process long and ambiguous statements, and ii) leading to a
“context collapse” where users lack contextual explanations and
justifications to aid their real-time privacy decision making (Vitak 2012).
Therefore, in raising users’ privacy awareness, it becomes imperative to
find an effective way to present and implement transparency.
In this article, we argue that privacy researchers who are interested in
addressing Big Data privacy challenges are likely to benefit from the
Fall 2015
80
emerging field of Human-Data Interaction (HDI) (Mortier et al. 2014). HDI
emphasizes on creating a collaborative but sometimes combative data
ecosystem around multiple stakeholders engaging in the collection and use
of personal data. The HDI approach does not throw out transparency
entirely, but gently refocuses this paradigm onto individuals’ privacy
awareness that would enable legibility , agency and negotiability.
Legibility is concerned with making data space (from collection, use,
analysis, to retention) both transparent and comprehensible (Mortier et al.
2014). To achieve the goal of legibility, researchers need to create
innovative mechanisms to visualize: who has collected what private data;
how the private data are being processed; how their private data are mingled
with others’ private data; what is done by the data brokers; and who are
using their private data and how. We argue that legibility empowerment is a
precursor to an individual’s ability to exercise agency in situations where
personal data are being collected and used.
Agency is concerned with giving people the capacity to act within the
data ecosystem. Consistent with Mortier and associates (2014), we do not
believe that all individuals should continually exercise this capacity; but
some of them can have the agency whenever they wish to. Thus privacy
researchers and technologists need to operationalize agency through
intelligent personalized approach by providing individuals with the
customized option of expressing concern over certain data use which they
do not agree with.
Negotiability is concerned with many dynamic relationships that
arise around data and data processing (Mortier et al. 2014). This theme
requires collaboration and engagement with stakeholders to collaboratively
decide what and why data exchanges occur, as well as specify the
information flow “from whom,” “to whom,” “for what reasons,” and “under
what conditions.” Placing discussions on negotiability empowerment
within the relevant contexts does not suggest that there is an agreement
about the level of privacy that is appropriate in any given context. However,
knowing the relevant dimensions and stakeholders of information flow in
the specific context does clarify the discussion.
Washington Academy of Sciences
81
Conclusion
This article suggests a new approach to conceptualizing privacy with
emphasis on emerging privacy threats in terms of information
identifiability, information ephemerality, and information linkability. The
latter two, in particular, are of growing importance, and pose significant and
unique threats to information privacy with the emergence and widespread
of Big Data technologies.
The reconceptualization of information privacy in these three
dimensions provides a unique opportunity for the emerging field of
Human-Data Interaction (HDI). It delineates three mechanisms through
which big social data analysis may influence users, and serves as a
theoretical foundation for future user-centered studies of privacy concerns
and privacy decision-making concerning Big Data practices and products.
This conceptual framework can further guide privacy research and ethics
discussions to draw economic, social and legislative implications of Big
Data practices, as well as finding practical solutions to these three privacy
challenges.
Acknowledgments
The authors gratefully acknowledge the financial support of the U.S.
National Science Foundation under grant CNS-0953749. Any opinion,
findings, and conclusions or recommendations expressed in this material
are those of the authors and do not necessarily reflect the views of the U.S.
National Science Foundation.
References
Acquisti, A. (2014). “The economics and behavioral economics of privacy”, in Privacy,
Big Data, and the Public Good: Frameworks for Engagement , :J. Lane, V. Stodden, S.
Bender, and H. Nissenbaum (eds) Cambridge: Cambridge University Press, 98-1 12.
Aldennan, E., and Kennedy, C. (1997). The Right to Privacy , New York: Vintage Books.
Ambrose, M. L. (2012). “It’s about time: Privacy, information lifecycles, and the right to
be forgotten”. Stanford Technology Law Review, 16, 101-154.
Fall 2015
82
Blanchette, J. F., and Johnson, D. G. (2002). Data retention and the panoptic society: The
social benefits of forgetfulness. The Information Society, 18(1), 33-45.
Brankovic, L., and Estivill-Castro, V. (1999). Privacy issues in knowledge discovery and
data mining. In Proceedings of Australian Institute of Computer Ethics Conference
(pp. 89-99).
boyd, d., and Crawford, K. (2012). Critical questions for big data: Provocations for a
cultural, technological, and scholarly phenomenon. Information, Communication &
Society, 15(5), 662-679.
boyd, d., and Ellison, N. B. (2007). Social network sites: Definition, history, and
scholarship. Journal of Computer-Mediated Communication, 13(1), 210-230.
Davenport, T. H., Barth, P., and Bean, R. (2013). How “Big Data” is different. MIT Sloan
Management Review, 54(1), 43-46.
de Wolf, R., Willaert, K., and Pierson, J. (2014). Managing privacy boundaries together:
Exploring individual and group privacy management strategies in Facebook.
Computers in Human Behavior, 35, 444-454.
Facebook. (2010). How happy are we? Accessed 30 October 2015:
https://www.facebook.com/notes/facebook/how-happy-are-we/ 150162112130.
Fogel, J., and[https://cups. cs.cmu.edu/soups/2009/proceedings/a4-kelley.pdf (accessed 30
October 2014)] Nehmad, E. (2009). Internet social network communities: Risk taking,
trust, and privacy concerns. Computers in Human Behavior, 25(1), 153-160.
Grudin, J. (2002). Group dynamics and ubiquitous computing. Communications of the
ACM, 45(12), 74-78. ’
Kelley, P. G., Bresee, J., Cranor, L. F., and Reeder, R. W. (2009). A nutrition label for
privacy. In Proceedings of the 5 th Symposium on Usable Privacy and Security.
[https://cups.cs.cmu.edu/soups/2009/proceedings/a4-kelley.pdf (accessed 30 October
2014)]
Knijnenburg, B. P. and Kobsa, A. (2013). Making decisions about privacy: information
disclosure in context-aware recommender systems. ACM Transactions on Interactive
Intelligent Systems, 3 (3), 20.
Kranzberg, M. (1986). “Technology and History: Kranzberg's Laws.” Technology’ and
Culture 27(3): 544-560.
Lin, J., Amini, S., Hong, J. I., Sadeh, N., Lindqvist, J., and Zhang, J. (2012). Expectation
and purpose: understanding users' mental models of mobile app privacy through
crowdsourcing. In Proceedings of the 2012 ACM Conference on Ubiquitous
Computing, (pp. 501-510).
Lampinen, A., Lehtinen, V., Lehmuskallio, A., and Tamminen, S. (201 1). “We're in it
together: interpersonal management of disclosure in social network services”. In
Proceedings of the SIGCHl Conference on Human Factors in Computing Systems (pp.
3217-3226).
Washington Academy of Sciences
83
Levine, Y. (2013). “What surveillance Valley knows about you”.
http://pando.com/2013/12/22/a-peek-into-surveillance-valley/ (accessed 30 October
2015)
Mayer-Schonberger, V. (2009). Delete: The Virtue of Forgetting in the Digital Age.
Princeton, N.J.: Princeton University Press.
McAfee, A., Brynjolfsson, E., Davenport, T. H., Patil, D. J., and Barton, D. (2012). Big
data: The management revolution. Harvard Business Review , 90(10), 61-67.
Mortier, R., Haddadi, H., Henderson, T., McAuley, D., and Crowcroft, J. (2014).
Human-Data Interaction: The Human Face of the Data-Driven Society. Available at
SSRN: http://ssrn.com/abstract=250805 1 (accessed 30 October 2015)
Narayanan, A., and Shmatikov, V. (2009). “De-anonymizing social networks”. In
Proceedings of 30th IEEE Symposium on Security and Privacy, (pp. 173-187).
Palen, L., and Dourish, P. (2003).” Unpacking privacy for a networked world”. In
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp.
129-136).
Pollach, I. (2006). “Privacy statements as a means of uncertainty reduction in WWW
interactions”. Journal of Organizational and End User Computing (JOEUC), 18(1),
23-49.
Richards, Neil M., and Jonathan H. King. (20 1 3). “Three Paradoxes of Big Data.” Stanford
Law Review Online 66 (41).
[http://www.stanfordlawreview.org/online/privacy-and-big-data/three-paradoxes-big
-data (accessed 30 October 2015]
Shein, E. (2013). “Ephemeral data”. Communications of the ACM, 56(9), 20-22.
Smith, M., Szongott, C., Henne, B., and von Voigt, G. (2012). “Big data privacy issues in
public social media”. In Proceedings of 6th IEEE International Conference on Digital
Ecosystems Technologies (DEST), (pp. 1-6).
Stutzman, F., Gross, R., and Acquisti, A. (2012). “Silent listeners: The evolution of privacy
and disclosure on facebook”. Journal of Privacy and Confidentiality, 4(2), 7-41.
Troshynski, E., Lee, C., and Dourish, P. (2008). “Accountabilities of presence: refraining
location-based systems”. In Proceedings of the SIGCHI Conference on Human
Factors in Computing Systems (pp. 487-496).
Viswanath, B., Mislove, A., Cha, M.,and Gummadi, K. P. (2009).” On the evolution of
user interaction in facebook”. In Proceedings of the 2nd ACM workshop on Online
Social Networks (pp. 37-42).
Vitak, J. (2012). The impact of context collapse and privacy on social network site
disclosures. Journal of Broadcasting & Electronic Media, 56, 4 (2012), 451-470.
Warren, S. D., and Brandeis, L. D. (1890). “The right to privacy”. Harvard Law Review,
4(5), 193-220.
Wigan, M. R., and Clarke, R. (2013). “Big data's big unintended consequences”.
Computer, 46(6), 46-53.
Fall 2015
84
White House. (2012). Consumer data privacy in a networked world: A framework for
protecting privacy and promoting innovation in the global digital economy.
Washington, DC: White House.
Wang, N., Grossklags, J., and Xu, H. (2013) “An online experiment of privacy
authorization dialogues for social applications”. In Proceedings of the 2013
conference on Computer supported cooperative work. (pp. 26 1 -272).
Xu, H., Teo, H.-H., Tan, B. C., and Agarwal, R. (2012).” Effects of Individual
Self-Protection, Industry Self-Regulation, and Government Regulation on Privacy
Concerns: A Study of Location-Based Service”s. Information Systems Research ,
23(4), 342-1363.
Xu, H. (2012). “Reframing Privacy 2.0 in Online Social Network”. University of
Pennsylvania Journal of Constitutional Law , 14(4), 1077-1 102.
Bios
Heng Xu is an associate professor of Information Sciences and Technology
at the Pennsylvania State University. Her current research focus is on the
interplay between social and technological issues associated with
information privacy. She has authored or coauthored over 100 research
papers on information privacy, security management, human-computer
interaction, and technology innovation adoption.
Haiyan Jia is a post-doctoral scholar at the Penn State University in the
College of Information Sciences and Technology. Her research interest
primarily focuses on the social and psychological effects of communication
technology ranging from Web to mobile apps to smart objects. Her current
work investigates online privacy in social and collective contexts.
Washington Academy of Sciences
85
Washington Academy of Sciences
1200 New York Avenue, NW
Room 1 1 3
Washington, DC 20005
Membership Application
Please fill in the blanks and send your application to the Washington
Academy of Sciences at the address above. We will contact you as soon as
your application has been reviewed by the Membership Committee. Thank
you for your interest in the Washington Academy of Sciences.
(Dr. Mrs. Mr. Ms.)
Business Address
Home Address
Email
Phone Cell
Please indicate:
Preferred mailing address
Type of membership
Business
Regular
Home
Student
Present Occupation or Professional Position
Please list memberships in scientific societies - and include office held:
Fall 2015
86
Journal of the Washington Academy of Sciences
Instructions to Authors
The Washington Academy of Sciences publishes its interdisciplinary
peer-reviewed Journal of the Washington Academy of Sciences four times a
year — Spring, Summer, Fall, and Winter.
1 . Deadlines for quarterly submissions are:
Spring - February 1 Fall - August 1
Summer - May 1 Winter - November 1
2. Draft manuscripts using a word processing program (such as
MSWord), not PDF.
3. Papers should be 6,000 words or fewer. With 7 or more graphics,
reduce the number of words by 500 for each graphic.
4. Include an abstract of 150-200 words.
5. Graphics must be in black & white or greytone. They must be
referenced in the text.
6. Use endnotes, not footnotes. The bibliography may be in a style
considered standard for the discipline or professional field represented
by the paper.
7. Submit papers as email attachments to the editor.
8. Include the author’s name, affiliation, and contact information —
including postal address. Membership in an Academy-affiliated
society may also be noted.
9. Manuscripts are peer reviewed and become the property of the
Washington Academy of Sciences.
10. There are no page charges.
Please see the Academy’s web site, www.washacadsci.org, for the library
subscription rate, listing of articles dating to 1899, and information on
accessing them.
Washington Academy of Sciences
87
Washington Academy of Sciences
Affiliated Institutions
National Institute for Standards & Technology (NIST)
Meadowlark Botanical Gardens
The John W. Kluge Center of the Library of Congress
Potomac Overlook Regional Park
Koshland Science Museum
American Registry of Pathology
Living Oceans Foundation
National Rural Electric Cooperative Association (NRECA)
Fall 2015
88
Delegates to the Washington Academy of Sciences
Representing Affiliated Scientific Societies
Acoustical Society of America, Washington Chapter
American Assoc, of Physics Teachers, Chesapeake Section
American Astronomical Society
American Ceramics Society
American Fisheries Society
American Institute of Aeronautics and Astronautics
American Meteorological Society, Washington, DC Chapter
American Nuclear Society, Washington DC Section
American Phytopathological Society, Potomac Division
American Society for Cybernetics
American Society for Metals, Washington Chapter
American Society of Civil Engineers, National Capital Section
American Society of Mechanical Engineers, Washington
Section
American Society of Microbiology, Washington Branch
American Society of Plant Biologists, Mid-Atlantic
Anthropological Society of Washington
ASM International
Association for Women in Science, DC Metropolitan Chapter
Association for Computing Machinery, DC Area Chapter
Association for Science, Technology, and Innovation
Association of Information Technology Professionals
Biological Society of Washington
Botanical Society of Washington
Capital Area Food Protection Association
Chemical Society of Washington
District of Columbia Institute of Chemists
District of Columbia Psychological Association
Eastern Sociological Society
Electrochemical Society, National Capital Section
Entomological Society of Washington
Geological Society of Washington
Historical Society of Washington DC
Human Factors and Ergonomics Society, Potomac Chapter
Institute of Electrical and Electronics Engineers, Northern
Virginia Section
(continued on next page)
Paul Arveson
Frank R. Haig, S. J.
Sethanne Howard
Vacant
Lee Benaka
David W. Brandt
Vacant
Charles Martin
Vacant
Stuart Umpleby
Vacant
Vacant
Daniel J. Vavrick
Vacant
Mark Holland
Vacant
Toni Marechaux
Jodi Wesemann
Alan Ford
F. Douglas
Witherspoon
Chuck Lowe
Stephen Gardiner
Chris Puttock
Keith Lempel
Elise Ann Brown
Vacant
Tony Jimenez
Ronald W.
Mandersheid
Vacant
Vacant
Jeffrey B. Plescia
Jurate Landwehr
Vacant
Gerald P. Krueger
Murty Polavarapu
Washington Academy of Sciences
Delegates to the Washington Academy of Sciences
Representing Affiliated Scientific Societies
(continued from previous page)
Institute of Electrical and Electronics Engineers, Washington
Section
Institute of Food Technologies, Washington DC Section
Institute of Industrial Engineers, National Capital Chapter
International Association for Dental Research, American
Section
International Society for the Systems Sciences
International Society of Automation, Baltimore Washington
Section
Instrument Society of America
Marine Technology Society
Maryland Native Plant Society
Mathematical Association of America, Maryland-District of
Columbia- Virginia Section
Medical Society of the District of Columbia
National Capital Area Skeptics
National Capital Astronomers
National Geographic Society
Optical Society of America, National Capital Section
Pest Science Society of America
Philosophical Society of Washington
Society for Experimental Biology and Medicine
Society of American Foresters, National Capital Society
Society of American Military Engineers, Washington DC
Post
Society of Manufacturing Engineers, Washington DC
Chapter
Society of Mining, Metallurgy, and Exploration, Inc.,
Washington DC Section
Soil and Water Conservation Society, National Capital
Chapter
Technology Transfer Society, Washington Area Chapter
Virginia Native Plant Society, Potowmack Chapter
Washington DC Chapter of the Institute for Operations
Research and the Management Sciences (WINFORMS)
Washington Evolutionary Systems Society
Washington History of Science Club
Washington Paint Technology Group
Washington Society of Engineers
Washington Society for the History of Medicine
Washington Statistical Society
World Future Society, National Capital Region Chapter
Richard Hill
Vacant
Neal F. Schmeidler
J. Terrell Hoffeld
Vacant
Vacant
Hank Hegner
Jake Sobin
Vacant
D. S. Joseph
Vacant
Vacant
Jay H. Miller
Vacant
James Cole
Vacant
Eugenie Mielczarek
Vacant
Daina Apple
Vacant
Vacant
E. Lee Bray
Terrell Erickson
Richard Leshuk
Vacant
Russell Wooten
Vacant
Albert G. Gluckman
Vacant
Alvin Reiner
Alain Touwaide
Michael P. Cohen
Jim Honig
Washington Academy of Sciences
Room 113
1200 New York Ave. NW
Washington, DC 20005
Return Postage Guaranteed
NONPROFIT ORG
US POSTAGE PAID
MERRIFIELDVA 22081
PERMIT# 888
4*8******”*1 1 9***«************AUT0**IVIIXED ADC 207
SERIAL RECORDS DIVISION
ERNST MAYR LIBRAY
MUSEUM COMP ZOOLOGY HARVARD UNIVERSITY
26 OXFORD ST
CAMBRIDGE MA 02138-2902
WAS
833^
Volume 101
Number 4
Winter 2015
Journal of the
WASHINGTON
ACADEMY OF SCIENCES
Board of Discipline Editors iv
JSTOR 1
Tribute to Julius "Jay" Earl Uhlaner 6. Kruger. 3
A Study of the Primary Granitoid Outcroppings and Sedimentary Rocks 7
N. Bassanganam, Yang Mei Zhen, Prince E. Y. Danguene, M. Wang
Candidates for the Source of the 1977 "WOW" Signal A. Paris, E. Davies 25
Affine Geometry, Planck Length and Cosmic Acceleration G. L Murphy 33
2015 membership list 45
Membership Application 53
Instruction to Authors 54
Affiliated Institutions 55
Affiliated Societies and Delegates 56
ISSN 0043-0439
Issued Quarterly at Washington DC
MCZ
LIBRARY
FEB 1 6 2018
HARVARD
UNIVERSITY
Washington Academy of Sciences
Founded in 1898
Board of Managers
Elected Officers
President
Mina Izadjoo
President Elect
Mike Coble
Treasurer
Ronald Hietala
Secretary
John Kaufhold
Vice President, Administration
Terry Longstreth
Vice President, Membership
Sue Cross
Vice President, Junior Academy
Vice President, Affiliated Societies
Gene Williams
Members at Large
Paul Arveson
Michael Cohen
Frank Haig, S.J.
Neal Schmeidler
Mary Snieckus
Past President
Terrell Erickson
Affiliated Society Delegates
Shown on back cover
Editor of the Journal
Sethanne Howard
Journal of the Washington Academy of
Sciences (ISSN 0043-0439)
Published by the Washington Academy of
Sciences
email: wasiournal(5)washacadsci.orq
website: www.washacadsci.org
The Journal of the Washington Academy
of Sciences
The Journal is the official organ of the
Academy. It publishes articles on science
policy, the history of science, critical reviews,
original science research, proceedings of
scholarly meetings of its Affiliated Societies,
and other items of interest to its members. It
is published quarterly. The last issue of the
year contains a directory of the current
membership of the Academy.
Subscription Rates
Members, fellows, and life members in good
standing receive the Journal free of charge.
Subscriptions are available on a calendar year
basis, payable in advance. Payment must be
made in US currency at the following rates.
US and Canada $30.00
Other Countries $35.00
Single Copies (when available) $1 5.00
Claims for Missing Issues
Claims must be received within 65 days of
mailing. Claims will not be allowed if non-
delivery was the result of failure to notify the
Academy of a change of address.
Notification of Change of Address
Address changes should be sent promptly to
the Academy Office. Notification should
contain both old and new addresses and zip
codes.
POSTMASTER:
Send address changes to WAS, Rm 113,
1200 New York Ave. NW
Washington, DC 20005
Academy Office
Washington Academy of Sciences
Room 113
1200 New York Ave. NW
Washington, DC 20005
Phone:(202) 326-8975
1 200 New York Ave.
Suite 113
Washington DC
20005
wwwwashacadsci.org
Journal of the
WASHINGTON ACADEMY OF SCIENCES
Volume 101 Number 4 Winter 2015
Contents
Editorial Remarks S. Howard ii
Board of Discipline Editors iii
JSTOR 1
Tribute to Julius “Jay” Earl Uhlaner G. Kruger 3
A Study of the Primary Granitoid Outcroppings and Sedimentary Rocks 7
N. Bassanganam, Yang Mei Zhen, Prince E. Y. Danguene, M. Wang
Candidates for the Source of the 1977 “WOW” Signal A. Paris, E. Davies 25
Affine Geometry, Planck Length and Cosmic Acceleration G. L. Murphy 33
Membership List for 2015 45
Membership List 53
Instructions to Authors 54
Affiliated Institutions 55
Affiliated Societies and Delegates 56
ISSN 0043-0439 Issued Quarterly at Washington DC
MCZ
LIBRARY
t i A K V ARD
UNIVERSITY
Winter 2015
ii
Editorial Comments
This Winter issue of the Journal has some interesting, eclectic things to
share. The Academy has signed a contract with JSTOR - a company that
digitizes science journals for online access. We start with a description of
that process. Eventually they will have digitized all our back issues.
Following that is a tribute to WAS Fellow Jay Uhlaner reprinted
with permission from the Human Factors & Ergonomics Society Bulletin
and written by G. Kruger.
Next is a graduate student paper that describes geological studies
done in the Central African Republic around the region of Boali. Noted for
its waterfalls Les Chutes de la Mbi is a 656-foot cascade where the Upper
M’poko River meets the Oubangui River. The natural beauty of the site has
earned it a place on the tentative UNESCO World Fleritage Site list. The
Falls of Boali are 250 m wide and 50 m high, and are a popular tourist
destination. I do not usually put photos in the editorial comments, but this
is an exception. The Falls of Boali are shown below.
Washington Academy of Sciences
iii
This is the first paper we have received from there. The author is a student
at the China University of Geosciences in Wuhan China. He speaks French
and Chinese but not English, so it was a challenge to get the paper first into
English and second into the structure of a technical and publishable paper.
It took several months of work, and I am proud to present the work for his
Master’s thesis.
Then comes a paper on the “wow” signal. This one has a back story.
In 1977 a radio telescope in Ohio received an intense, short signal. At that
time data were recorded on a chart recorder - a bit like a lie detector setup.
A sheet of paper rolls out with a moving pen recording the data. The signal
the telescope received was so large that the telescope operator wrote “wow”
on the paper - hence the name the “wow” signal. Talk to any radio
astronomer, mention the “wow” signal, and they will know the story. To
date it has not been explained. This paper offers a possibility for the signal.
Last up is a paper that discusses the metric based general theory of
relativity and dark energy. The author provides a view of the problem of
dark energy based on Schrodinger’s affine field theory. Put on your serious
math hats for this one. It may explain that odd “cosmological constant”, A.
As usual the Winter issue ends with a list of members of the
Academy.
Editor
Sethanne Howard
Winter 2015
IV
Journal of the Washington Academy of Sciences
Editor Sethanne Howard sethanneh@msn.com
Board of Discipline Editors
The Journal of the Washington Academy of Sciences has a 12-member
Board of Discipline Editors representing many scientific and technical
fields. The members of the Board of Discipline Editors are affiliated with a
variety of scientific institutions in the Washington area and beyond —
government agencies such as the National Institute of Standards and
Technology (NIST); universities such as Georgetown; and professional
associations such as the Institute of Electrical and Electronics Engineers
(IEEE).
Washington Academy of Sciences
The Journal of the Washington Academy of Sciences joins
the JSTOR Archive
i
The Washington Academy of Sciences has signed an agreement
with the JSTOR archive dedicated to preserving scholarly literature. The
complete back run of the Journal of the Washington Academy of Sciences
(JWAS), which dates to 1899, will be digitized and made available via the
JSTOR online platform.
In addition to the Washington Academy, more than 1,050
publishers, including scholarly societies and publishing academies of
sciences, are currently part of the JSTOR archive which hosts some 2,200
digitized journals comprised of 9 million digitized articles in various
collections. For example, the oldest journal in the JSTOR collections is the
Proceedings and Transactions of the Royal Society of London , which
dates back to 1665.
More than 8,000 institutions from 175 countries make use of
JSTOR, including universities, secondary schools, government and non-
profit organizations, community colleges, museums, and public libraries.
The Academy’s former president Terrell Erickson says, “This is a terrific
opportunity for the Washington Academy of Sciences as it expands our
Journal’s reach beyond our current subscriber base to a much larger
audience.”
Several programs help to make sure that the archive’s contents are
widely-available at a reasonable cost to users such as students and other
science professionals. For instance, in addition to the 8,000+ subscribing
libraries, JSTOR is also available to individual unaffiliated researchers
who can access single articles through various JSTOR accessibility
programs like “Register & Read” and “JPASS.” JSTOR further supports
the African Access Initiative (AAI) and Developing Nations Access
Initiative (DNAI), and these initiatives waive or reduce fees for 1,279 not-
for-profit and academic institutions in developing countries.
The not-for-profit JSTOR archive was conceived to help libraries
and publishers respond to the rising costs associated with the storage of
printed journal literature and to ensure that this material would not be
“lost” as academic research became increasingly electronic. Through the
digitization of complete journal runs, JSTOR makes it possible for
Winter 2015
2
subscribing libraries to share the costs associated with storage and
maintenance of journal literature, as the non-destructive digitization
process will be done at no cost to the Academy. Furthermore, JSTOR is
offering the Academy a modest revenue-sharing arrangement based upon
access to JSTOR by its users.
JSTOR will work from print copies of JWAS to create image files
that are exact replicas of the original Journal pages and text files that
enable searching. Upon completion of this process, users will be able to
conduct full-text searches back to the first volume and issue in 1 899 when
JWAS was called the Washington Academy of Sciences Proceedings.
Scholars will then be able to browse, search, view, and print JWAS and the
earlier Proceedings directly from their desktops.
The Academy will retain the copyright to the material published in
its Journal , as the JSTOR license agreement is non-exclusive. The
Academy is planning to convert the current JWAS hard-copy format to an
online version, and is exploring hosting JWAS electronically at its own
website for its members and numerous paid individual and institutional
subscribers who will be the only ones to have access the current and recent
issues of JWAS. There will be a 3-year gap between the most-recently
published issue of the Journal and the last issue available in JSTOR. This
window of time is being designated for the puipose of separating these
paid subscribers and members from the older issues which will be
available via the archive.
Questions about the Academy’s Journal can be directed to JWAS
editor Sethanne Howard, sethanneh@msn.com .
Washington Academy of Sciences
3
MEMBER MILESTONES
Julius “Jay” Earl Uhlaner ( 1 9 1 7 - 20 1 5)'
By Gerald P. Krueger
Human Factors and Ergonomics Society (HFES) Fellow
Julius “Jay” Earl Uhlaner was born in Vienna, Austria,
in 1917. In 1928, he immigrated to the United States,
where he became a naturalized citizen and left a lasting
legacy through his leadership and research
achievements, especially in applying psychology to
military problems.
, .. „ , .... Jay graduated from City College of New York in 1938
Julius Earl Uhlaner
with a BS in science. He worked in human engineering
at Ford Motor Company in Michigan from 1939 to 1940 and established a
driver research lab. In his early human factors work, he focused on driver
vision, training, and safety issues. These interests led to his thesis work for
his MS in psychology and statistics from Iowa State University in 1941 . His
contributions to highway safety included significant research on the
visibility and interpretability of roadway signs with different types of
lettering (e.g., height/width ratios of letters). He served on the Highway
Safety Research Board in Lansing, Michigan, and dealt with human factors
issues.
While serving as a psychologist in the Army Air Coips during
World War II from 1943 to 1946, Jay was involved with developing criteria
for selecting pilots. From 1946 to 1947, he was assistant director for
research and training for the New York State Division of Veteran Affairs.
Combining his bent for human factors and personnel selection, he earned a
PhD in industrial and organizational psychology at New York University in
1947.
Jay then joined the Army Personnel Research Branch as a research
psychologist. As the organization grew, it eventually became the Behavior
Systems Research Lab (BSRL). In 1969, Jay became BSRL technical
1 Human Factors & Ergonomics Society Bulletin , 58. No. 10, October 2015
Winter 2015
4
director. Two years later, he also took on the title of chief psychologist of
the U.S. Army, which is still worn today by the director of BSRL’s even
broader-based successor organization: The Army Research Institute (ARI)
for the Behavioral and Social Sciences. Under Uhlaner’s visionary
guidance, ARI gradually took on missions to develop and improve the
performance of people in the Army through behavioral sciences research on
personnel selection, classification, job placement, training systems, and
human factors in systems design. With Uhlaner at its helm from 1969 to
1978, ARI grew to employ more than 400 research psychologists, many of
them well steeped in and practicing classical human factors methods and
attaining many noteworthy accomplishments.
Jay was best known for some of his innovative contributions to the
Army. He foresaw early on the movement toward reliance on computers
and automation and had ARI focus on “person-in-the-loop” approaches to
examining soldier-system interface situations wherein the infusion of new
technologies could enhance soldier performance, training systems, and
equipment system testing. He spearheaded development of the first
psychological military qualifications test legislated by Congress; introduced
computers as major tools and partners in behavioral science research;
pioneered research on night-vision testing and driver performance;
introduced the first classification system based on psychological aptitude
testing in the military services; pioneered the “system measurement bed,” a
methodology that influenced industrial psychology; and fostered an
interdisciplinary approach to ARBs research.
During his career, Jay
published close to 200 articles in
scientific journals and books on the
subjects of industrial psychology,
military psychology, and related
topics. In 1976, President Gerald R.
Ford awarded him the U.S.
Presidential Award for Management
Improvement for his commanding
role in the development and
implementation of the Army
Classification Battery and Aptitude
Jay Ulilaner as director of the U.S. Army
Behavior & System Research Laboratory
Area System, representing major
Washington Academy of Sciences
5
advances in the field of soldier performance prediction. In 1995, the
American Psychological Association’s (APA) Division 19 (Military
Psychology) recognized Jay with the Lifetime Achievement Award in
Military Psychology for his many accomplishments in the application of
behavioral science research to military problems. In 2011, Division 19
initiated an award in his name: the Julius E. Uhlaner Award for
Distinguished Contributions to Research on Military Selection and
Recruitment.
In addition, Jay was a Fellow of FIFES, APA, and the Washington
Academy of Sciences (WAS). In 1976, WAS granted him the first award
“for scientific work of high merit in behavioral sciences” (see below).
After retiring from the Army in 1978, Jay was senior vice president
at Perceptronics, Inc., a human performance modeling, simulation, and
training consulting firm in California (at that time). One of the more notable
programs he fostered as part of a consortium for Defense Advanced
Research Projects Agency (DARPA) was SIMNET, which offered a tank
battle 3-D virtual simulation training network that permitted dozens, if not
hundreds of operators in tanks, helicopters, close support aircraft, and other
battlefield entities to interact with one another during war game training. At
Perceptronics, Jay also did extensive work in mining safety for the
Department of Commerce. He retired in 2000 but continued as a member of
the board of directors. Subsequently, he carried out his own part-time
behavioral sciences consulting work for another decade.
Having watched him from a short distance, I can say that Jay
Uhlaner continually demonstrated significant political and scientific savvy
in dealing with bureaucracy and in getting things done. He was particularly
adept at obtaining buy-in to build up human factors research psychology in
the military by having his staff seek to provide what the country’s leaders
and soldiers needed most.
Jay’s family can be contacted through his beloved wife of 66 years,
Vera Uhlaner, at P.O. Box 967, Corona del Mar, CA 92625-9998.
Winter 2015
6
Award by the Washington Academy of Sciences
The presentation was made at the Annual Awards Dinner meeting of the
Academy on Thursday, March 18, 1976, at the Cosmos Club.
Dr. Julius E. Uhlaner, Chief Psychologist of the U. S. Army and Technical
Director of the Army Research Institute for the Behavioral and Social
Sciences, and Adjunct Professor of Psychology at George Washington
University, was cited for “his outstanding technical direction and leadership
in Applied Psychology.” As a psychologist, he is best known for
contributions to military psychology, having spent the major part of his
career as a civilian research psychologist in the Army. However, he also
kept closely in touch with academia and industry. He is best known for some
of his innovative contributions to the Army, having developed the first
psychological military qualifications test legislated by Congress; introduced
the use of the computer as a major tool and partner in Behavioral Science
research; pioneered night vision testing research and driver research;
introduced the first differential classification system based on psychological
aptitude testing anywhere in the military services; pioneered the “system
measurement bed,” a methodology which influenced the field of industrial
psychology; and fostered the interdisciplinary approach to much of his
research. Also, he has exhibited very active professionalism, including the
holding of elective offices in divisions of the American Psychological
Association.
His awards in the Federal service include the Citation for Meritorious
Civilian Service, 1960; Citation for Exceptional Civilian Service, 1969; and
Citation for Outstanding Performance, 1972.
His combination of experience and education led to his trademark for the
conduct of research in the Behavioral Sciences — an interdisciplinary
approach, systems oriented, and the use of research products.
He was elected a Fellow of the Washington Academy of Sciences in 1963;
he was also a Fellow of the American Psychological Association. He was a
Fellow of the Human Factors Society and the Iowa Academy of Sciences.
Other societies of which he is a member are the Operations Research
Society of America, International Association of Applied Psychology,
Psychonomics Society, and District of Columbia Psychological
Association.
Washington Academy of Sciences
A Study of the Primary Granitoid Outcroppings and
Sedimentary Rocks in the Boali Region of the Central
African Republic
7
Narcisse Bassanganam, Yang Mei Zhen, Prince E. Yedidya
Danguene, Minfang Wang
Earth Resource, China University of Geosciences, Wuhan, China
Earth Faculty, University of Bangui, Central African Republic
Abstract
Located in the Central African Republic, the region of Boali is noted for its
waterfalls and for the nearby hydroelectric projects. The waterfalls of Boali
are 250 m wide and 50 m high, and are a popular tourist destination. The
Central African Republic (CAR) has large reserves of Granitoids that remain
largely untapped. That is why these rocks, which outcrop and which constitute
the base of the Boali region and its surroundings, caught our attention.
Previous studies by Bowen (1915) explained the order of appearance of
various minerals as a function of the temperature and initial magma (SiCE)
content. According to Bowen’s diagram, we can say that the magma
underwent a magmatic differentiation giving rocks that are poor in silica
(Diorite), followed by rocks rich in silica (Granodiorite and Granite). Knowing
the absolute age of the Granitoids on the edge of the craton of Mbomou (2.1
Ga, Moloto et al., 2008, and Toteu et al, 1994), we can deduce the chronology
of other formations. Initially there was the formation of the metamorphic
formations and sandstones of Boali. This was followed by a slow intrusion of
magma which crystallized in depth to give grainy rock (granitoids and
pegmatite) in the region of Boali. This intrusion had metamorphosed the pre-
existing formations through an orthogneiss.
Introduction
Boali is a town located in the Ombella M’poko prefecture of the Central
African Republic (CAR) (See Figure 1). It is located 100 km northwest of
Bangui, the capital of CAR. Boali is between 18°7'0"E longitude, and
4°48'0"N latitude. Access is through the National Road 1 (RNl).
Boali is a sub-prefecture in the CAR. The CAR is divided into 16
administrative prefectures, two of which are economic prefectures, and one
an autonomous commune; the prefectures are further divided into 71 sub-
prefectures. The prefectures are Bamingui-Bangoran, Basse-Kotto, Haute-
Kotto, Haut-Mbomou, Kemo, Lobaye, Mambere-Kadei', Mbomou, Nana-
Winter 2015
8
Mambere, Ombella-M'Poko, Ouaka, Ouham, Ouham-Pende and Vakaga.
The economic prefectures are Nana-Grebizi and Sangha-Mbaere, while the
commune is the capital city of Bangui.
Fig. 1: Map of CAR showing Ombella M’poko and Boali study area
The Central African Republic is a country rich in mineral resources
with an important reserve of Granitoids. Granitoid or granitic rock is a
variety of coarse grained plutonic rock similar to granite which is composed
predominantly of feldspar and quartz. These rocks outcropped and
constitute the base of the Boali region, but unfortunately are not exploited.
Geologically Boali is very interesting because of its Granitoids. We
will identify and define the importance and usefulness of the Granitoids not
only to geology, but also for the economy and social development in the
CAR. We note that a school was built at Crossing-Boali in 1953 by the priest
Alosiste Gezst, and recently, in 2001, the College of General Education
(C.G.E) was built.
Washington Academy of Sciences
9
Early geologic studies by Bowen (1915) defined the order of
appearance of various minerals. Cornacchia et al. (1989) described the
geologic formations of Boali. They mentioned that the greenstone belt of
Bogoin-Boali rocks represents a succession of structures with a narrow
synclinal appearance drawing a large half circle. These structures end in the
east in the Bogoin area and to the north in the Boali sector as the
outcroppings observed north of Boali.
Poidevin (1979) defined the geochemistry of Precambrian basaltic
rocks from the CAR; at Mbi not far from the river M’poko there are three
types of petrographs: Schist sencitic, chlorite schist, and quartzite.
Cornacchia and Dars (1983) showed that a corridor of faults cut north of the
CAR existed. Cornacchia et al. (1985) found in the sandstone quartz veins
containing crystals of rocks. Poidevin and Pin (1986) showed that the
outcropping is plural-kilometric with an intrusion of dolerite and granites.
Lithological studies of the Boali-Bogoin-Mbi region by Cornacchia
and Giorgi (1986) defined a vast area ranging from the border of the
republic of the Congo to south of the Lobaye Subit-Possel road including
the Boda area. Their work was earned out south of the M’poko River and
continued from the town of Bogoin to Yangana up to the Yasi series in the
area of Bangui.
Biandja (1988) earned out his work largely in the northern region of
the Bogoin. Biandja (2000) pointed out that the southern part of the Boali
region is characterized by a series of “Mbi” (waterfalls) incorporated from
the bottom upwards. The series contains amphibolites of Mbali and Mbi and
pillow basalts. All the intruded granite is in the lower course of the river
Mandjo. North of the Bako village on the Mbi, this succession of granite
becomes abnormal when it contacts the red sandstone and the red shale of
the base of the sandstone shale set. In the northern region of the Bogoin
there is a succession of chloritized migmatite and amphibolites that include
some biotite in the faults area. There is also migmatized ferruginous
quartzite. The sub-horizontal schistose sandstone does not conform to the
christallophylliennes formations. However, the whole region of Boali does
show some similarities between the north and south.
According to studies done by Poidevin (1991), Biandja (1988); and
Cornacchia et al. (1985-1989), the Boali region forms the southern part of
a greenstone belt that represents the northwestern part of Bogoin-Boali. The
Winter 2015
10
orientation of this greenstone belt runs east-west ending at the eastern edge
and marries with an intrusive granite border at the western edge. According
to the report from the meteorological station of Mbali covering 1993 to
2000, the average annual rainfall generally ranges from 1900 mm in
February to 2630 mm in December, with an average maximum of 2868 mm.
The Granitoids are on the edge of the craton of Mbomou (age 2.1 Ga)
(Moloto et a/., 2008; Toteu et a/, 1994). A craton is a large, stable block of
the Earth’s crust forming the nucleus of a continent. Recent studies by Rolin
(1992) focused in the Central African Republic area of pan-African strike-
slip of the Oubanguides. In general, Djebebe-Ndjiguim (2013) found that
the density of the vegetation made it very difficult to search for significant
outcrops.
We continue their work to include not only new information on the
geologic formation of the Boali region, but also to note the effect that non-
exploitation of the granitoids in the area has on the region. It is a complex
issue. Consequently the granitoids have not contributed to the social
development in the Boali area in particular and to the CAR in general.
Techniques Used to Gather the Data
Boali is located 100 km northwest of Bangui, the capital of CAR.
This field study was done on 24/25 June 2015. We used the basic tools of
the geologist: a compass, camera, hammer, bag, notebook, and pencil. Out-
general approach is based on the work of Cornacchia and Giorgi (1986). As
noted by Djebebe-Ndjiguim (2013) the amount and density of the
vegetation made it very difficult to search for significant outcrops. The
authors followed two protocols set by previous researchers.
The first protocol we followed was that of Biandja (1988). His work
was carried out largely in the Bogoin northern region. In his lithological
description he was able to list petrographic features consisting of lateritic,
ferruginous, and conglomeratic blocks for recent formations. They
contained, on average, quartzite, white quartzite, sandstone quartzite for
covering the proterozoic formation; meta-volcano sedimentary, ferruginous
quartzite, gneiss, Amphibolites, meta-volcanic basic to ultra-basic schist,
and Metarhyolitoids (meta-volcanic acid) for the base formations of
metamorphic rocks. For the intrusions Biandja distinguished many
Washington Academy of Sciences
characteristics of crystalline Granitoid intruding porphyroides granites from
the base.
The second protocol we followed was that of Poidevin (1991). His
work was also earned out north of Bogoin. He identified different
petrographic characteristics and classified them by stratigraphic unit (as U/?
where n is a number 1 to 4). His four classifications are: Andesite in pillow-
lavas and chlorite amphibolites for the main basalt unit (Ul); Para
amphibolites, meta-rhyolites, with greywacke, feldspathic quartzite to
amphiboles for the intermediate unit (U2); the greenstone and many pillow-
lavas for the upper unit (U3); and Itabirite (U4). In addition to his four
stratigraphic units, he also revealed the existence of geological formations
of regional importance such as the granitoids and the series of schisto-
quartzitic rocks.
The Geologic Data
We studied a variety of rocks types: plutonic; sedimentary;
metamorphic; and deformations of rocks. In general, the extent and density
of the local vegetation made it very difficult to search for significant
outcroppings. (Djebebe-Ndjiguim 2013). We will consider the variety of
rocks type by type.
Plutonic rocks
A pluton is a body of intrusive igneous rock (called plutonic rock)
that is crystallized from magma slowly cooling below the surface of the
Earth. In this category we studied two types: quartz veins and Granitoids.
Quartz vein (lode): There are two types of quartz veins in the study sector:
metamorphic formations; and rock crystal veins located in the sandstone.
Quartz veins are not barren of mineralized rock crystals. And so in these
veins we noted the presence of some minerals, such as emerald and gold,
due to the movement of warm waters (Comacchia et al. 1985).
In the greenstone belt toward the vein wall there are altered
Amphibolites in the chlorite-schist. According to Cornacchia et al. (1985),
quartz veins containing rock crystals are found in the sandstone. These veins
continue through to the quartz veins found in the metamorphic rocks. They
originate in the emanation from granite and are mineralized rock crystal.
The veins occur from 30° N to the south with a thickness of 15 centimeters
Winter 2015
12
to 5 meters and orient North 90° to 105°. They include some geodes in
which beautiful quartz crystals have developed. The crystals have a
thickness of at least 30 millimeters. They can reach 1.5 m thick in some
geodes. Across the rock outcroppings in the direction N 30° they form solid
blocks of milky-white appearance and are poly fractured.
During our field observations we spotted four levels of
implementation veins in the quartz downstream of the dam at the Mbi, and
even more implementation veins next to the road to Bossemmbele. These
are the extension of those downstream of the dam. The seams are flush to
both sides of the hill overlooking the dam. Some veins fold into a semicircle
under the mast of the town’s police station and also in the stone quarry.
Granitoids: Granitoids are plutonic rocks that are poor in silicon dioxide
(SiCh). They are designated in the upper part of the table of the international
classification of streckeizen. In our region of Boali there are diorites,
granodiorites, granites of Mbi, and granites of Bolen. We observed that
granitoid outcroppings in the region cover a very large area. Although grey
in appearance these rocks sometimes have alternating beds of dark
ferromagnesian amphibole and biotite and clear beds (quartz, feldspar, and
muscovite).
The granitoids of Mbi orient 30° N dipping 70° W. They are
traversed by quartz aplite and pegmatite veins. These formations are
subdivided into granite, matching granite, and orthogneiss.
Granite of Mbi - Granite is a fully crystalline rock. Minerals are on average
2 to 5 mm in size about the size of a grain of wheat (granite comes from the
Latin granurn = grain). They contain three essential minerals: quartz, alkali
feldspar (orthoclase and microcline), and plagioclase combined with mica
(biotite and muscovite). The quartz comes in a grayish color surrounding
other crystals. Its appearance is that of salt but with a bold loamy appearance
as if it burst out of the rock. In the region of Mbi the quartz has a conchoidal
fracture. The alkali feldspars have variable colors (white, pink, red) and are
twin Karlsbad (the crystal is alternately brilliant and dull). Biotite occurs in
black strips some with a golden luster, with cleavages or cleavage lines.
This intrusive massif has lagged behind the plate tectonics. It is late granite,
very marked, and located on the left bank of Mbali in our study area. This
massif has a grainy central facies with large elements of alkaline feldspar,
very rich in feldspar, and with very fine grain borders. It manifests itself in
Washington Academy of Sciences
13
the landscape by significant outcroppings and can be observed in the stone
quarry without major difficulty.
Granodiorites: Granodiorites have a constitution nearly that of granite;
their silica content can be as strong as that in granite but contains more
plagioclase feldspar than orthoclase feldspar. Common rock “granite” can
be distinguished from granodiorites by carefully considering their feldspar.
Granodiorites of the Boali region have micro-fractures that allow the
circulation of fluids. There is a possibility of finding gold and pyrite. The
presence of the epidote gives the rock its green color. This epidotisation is
due to the alteration of the potassium in the feldspar. The outcropping is
plural-kilometric with an intrusion of the dolerite and granites. They are
dated to 2.1 Ga. (Poidevin and Pin 1986)
Diorites: Diorite is an intrusive igneous grainy rock with a silica deficiency
(less than 20%); therefore, it does not contain free quartz. It is principally
composed of the minerals plagioclase feldspar (typically andesine), biotite,
hornblende, and/or pyroxene. Feldspar, generally grayish, helps to give the
rock a dark color. Diorites are intruding amphibolites and are contiguous
with the granodiorites of Mbi.
Dolerite: Dolerites are intermediate rocks that fall between grainy gabbros
and basalts with microlitic grain that is visible under a microscope and
shows sub-hedral plagioclase laths molded by interstitial pyroxene. They
are generally massive and compact with a color ranging from black to grey
but more often dark green. We saw three hills of dolerite in the intruding
granodiorites in a nearby outcropping of granite.
Contact areas: The Bangui-Boali section shows several contact areas
characterized by vein crates between sedimentary rocks and metamorphic
rocks. About 1 00 km from the first dam to the north of Boali we find a
contact between Amphibolites and sandstones. The contact is characterized
by a puckered quartz hill. At 123 km another contact is characterized by a
type of vein that is favored by the hydrothermalism between granitoids and
massive Amphibolites; there is another contact with upright schistosity
(sub-vertical N 60°).
Winter 2015
14
Sedimentary rocks
Sedimentary rocks are rocks that are formed by the deposition of
material at the Earth’s surface and within bodies of water. Sedimentation is
the collective name for processes that cause mineral and/or organic particles
(detritus) to settle and accumulate or minerals to precipitate from a solution.
Sandstone is a sedimentary rock. It is a consolidated rock that belongs to
the class of arenite rocks that have a grain size between 0.0625 and 1 mm.
Thus we can distinguish between quartz sandstones, where a
microcrystalline material persists between the grains of quartz, and quartzite
sandstone, where grains are linked to each other following a secondary
pathway that depends on the cement. They are located south of the
Kassango area and belong to the Oolitic sandstone of the Boali series (see
Photo 1C which shows the sandstones of Boali falls). The corresponding
features are homogeneous fine-grained quartzose sandstones (with clay
cement in the south-west that changes to siliceous cement to the south and
south-east). Quartzite occurs on the road to the city in the stratigraphic
extension falls downstream from the third Boali falls. The sandstones are
grayish and friable rocks whose diamante detrital minerals are amorphous
quartz grains often recrystallized as anhedral feldspar. Observation with a
microscope reveals rare biotite lamellae and a few fine flakes of muscovite.
Boali sandstones are the equivalent of those of Fatima, a district located in
the Bangui capital of the CAR.
Metamorphic rocks
Metamorphic rocks arise from the transformation of existing rocks,
in a process called metamorphism, which means “change in form”. We
found four types of metamorphic rocks in our study area: Schist,
Amphibolites, Itabirite, and Gneiss.
Schist: Schists are characterized by medium to large, flat, sheet-like
grains with a preferred orientation. The outcroppings form in slabs on the
bed of the Kassango at the roadside and are often interstratified with the
sandstones. Of greenish hue this rock has a sub-vacuolar structure
throughout. It presents numerous vacuoles and therefore it is strongly
schistose. It fits into beds that are clear of recrystallized quartz and chlorite
and has dark beds of rare sericite altered biotite. We find these in the region
of Boali, and we find this same shale on the road to Damara. These are rich
Washington Academy of Sciences
15
in mica and nodules and are very crumpled; this is the schist of Boali, the
equivalent of the Fatima shale that belongs to the series of Bangui, which
are above the Yangana shale.
Photo 1 : sandstones of Boali falls
Amphibolites: Amphibolites are dark green rocks consisting mainly of
amphibole crystals more or less ordered along the planes of schistosity. We
can distinguish laterized amphibolites, layered amphibolites, and massive
amphibolites.
Highly altered and chloritized laterized amphibolites are found in
the area of the lakes of the crocodiles in a stone quarry about 100 km away
Winter 2015
16
from the laterites on the highway. Layered amphibolites are amphibolites
that have a banded texture characterized by alternating feldspathic quartz
beds and detachment beds. The hyper-fractured veins come within 1 km of
the Bogoin village (village Bobissa). Massive amphibolites are mottled with
Granoblastic massive rocks. The fine-grained rocks show a discontinuity in
their arrangement. The dark minerals are dominant with a cleavage of
amphibole and biotite. There are also some rare glitters of muscovite. The
massive amphibolites stretch from the Bogoin village to where they make
contact with the granodiorites.
Itabirite : The itabirites are quartzite ferruginous rubane. The outcropping is
in a kilometer wide band. These are generally quartz-rich rocks occurring
with magnetite and often oligiste. This last mineral concentrates in massive
structures around the quartz veins crossing the Banded Iron Formation
(BIF) where the banding is very marked; there is a layout of dark magnetite
beds alternating with beds of clear quartzo-feldsparthic rock.
Gneiss: The gneisses are medium-grained or coarse rocks about 1 mm to 1
cm in size. They often have net foliation characterized by beds of generally
dark hue, rich in minerals (mica, amphibole), and alternate with clear beds
of ferromagnesian (white grey, pink) quartz and feldspar visible to the
naked eye. We noted the presence of the orthogneisses, which are rocks that
form a contact between amphibolites and granodiorites on one side and form
a contact between the granite and granodiorites on the other side.
Other metamorphic rocks
The quartzite and muscovites rocks occupy the eastern part of the
region. Shale appears in slabs on the bed of the Ngalou. Chloritoschistes
and schist outcroppings occur in the region of Bogoin. Orthogneisses
occupy the southern part of the region of Bogoin. The southern region of
the Kolango is characterized by a lower relief that is very soft sided in its
uppermost part. On the sides of the rocky massive benches, block elements,
and fractured outcroppings we can distinguish massive metabasalts in the
pillow lavas and metabasalts in the stringers of intruding quartz.
Washington Academy of Sciences
17
Deformations
Deformation takes an object from its initial state to its final state by
mass transport (translation, displacement, rotation, and by internal
deformation). The deformed object is defined by its dimensions.
Stratification is one form of deformation. Bedding planes illustrate the style
of the planar structural element. These were initially roughly flat, horizontal
surfaces. Their characteristics and variations are an imprint of deformations
that have been imposed by the sedimentary terrain since their deposition.
This stratification is observed in the sandstone outcropping. At the entrance
to the falls of Boali we observed the stratification cross the sandstone (See
Photo IE).
Geological foliation (metamorphic arrangement in layers) with
medium to large grained flakes in a preferred sheetlike or planar orientation
is called schistosity. The plane of the schistosity is called S. In formations
containing more competent levels, stretching leads to socking which is to
say leads to the segmentation of the most competent object into fragments
and socks. Photo ID illustrates deformations characterized by boudinage,
folds, faults, and shears.
• Boudinage is a term used in geology to indicate structures formed
by extension (where a rigid body is deformed often into a sausage
or boudin like shape).
• A fold is a permanent waveform deformation in layered rock (the
rocks bend or twist). It occurs when one or a stack of originally flat
surfaces (such as sedimentary rock) are permanently bent or curved.
• A fault is a fracture in the bedrock. They are breaks accompanied by
the relative movement of two components. The movement can be
vertical (vertical, oblique, fault normal or reverse) or horizontal
(strike-slip or shear).
• A shear is the response of a rock to deformation usually by
compression. The shear can be emphasized by certain minerals.
We essentially observed the schist as shears and lineaments. These
are break planes that are accompanied by the relative movement of two
components which show the hang of the faults. Lineaments are mineral
lineations that occur during metamorphic crystallization.
Winter 2015
18
In our study area there is a boudin fold ply spilled to the NW. It
consists essentially of anhedral quartz crystals a centimeter in size. It is
molded into the clay schist (shale) and found at the roadside in Kassango.
Quartz flanges are located in the clay schist of Kassango. They can be found
at the level of the boudin folds. On the road to the town’s police station we
find a crease spilled in sandstone. The itabirite are also very creased. The
wrinkles are crooked with a very upright fold axis. Under the mast of the
police station is a surrounding concentric fold with a diameter of 40 cm.
The fold shown in photo ID is observed in the clay schist (shale) of
Kassango. Finally, we see a deformation characterized by a fold slumping
downward. Accompanying the schistosity and the boudin is a tangential
tectonic surface with direction S-SE toward N-NW. A second deformation
is a tangential tectonic surface contrary to the first. It runs NW-SE. This
tectonic surface is confined by the mega fold conic running N-NW. A
tectonic surface relates to the structure of the Earth’s crust and the large-
scale processes which take place within it.
Shears : Sinistral and dextral shears were observed at the stone quarry (S2,
See Figure 2). They form a corridor of sinistral shear 5 m wide for the S2
shears and fall 155-45° SW. The basal formation of the stone quarry shows
deformation bands approximately 60 m wide. We found that the dextral
shear (S2, Figure. 2) was hardly observable. On the other hand, the S2
shears are very representative of the class.
Figure 2: The center insert shows the dextral Shears in the area. The left
and right sides of the figure show the sinistral shear, S2
Washington Academy of Sciences
19
Lineaments: The lineaments are mineral lineations during metamorphic
crystallization. We observed lineaments in the sandstone towards the falls.
The lineament runs N 120-35° S. It crosses all fractures. We observed in the
dolerites two families of lineaments: one in direction N 135° sub-vertical;
and the other, south-facing N 25° with a dip of 60°. There are small
intercalations of gneiss in the outcroppings of dolerite. The thickness of
these dolerites can reach 40 m. Diabase dykes continue to the top of the hill.
All these formations in the sector are affected by brittle deformation which
appears here as faults. The faults are the fractures in the bedrock. They are
breaks accompanied by the relative movement of two components. The
movement can be vertical (vertical, oblique, fault normal or reverse) or
horizontal (strike-slip or shear). The fault of Boali is a normal fault (see
Photo 1 E. F) corresponding to Figure 3, which shows a normal fault. These
faults have been found in the sandstone in front of the police station (in the
main city). They include three (3) series of fracturing. FI and F2 in direction
N 0+/-100 they have an embedding of 60 -75° E, sub-vertical; and F3 in
direction N 145 +/- 10 they are sub-verticals. See Table 1 (in front of the
Internet service provider for Boali, in the main city).
In addition, the brittle tectonic of the Mbi sector highlights four
major series of faults: These series of faults run: N45° - N 50°; N 80° - N
100°;N130°-N 140°, corresponding to F3; and N160° -N 175° (see Table
1 (Mbi sector)).
Fig. 3: the shear dextral and normal fault corresponding to the study areas
observed
Winter 2015
20
Table 1: faults family observed in the study areas.
Photo 2: The basic faults and veins
Timeline for the Faults: Photo 2 shows the various fault lines we observed.
First FI was established with a filling. Then a second fault F2 parallel to FI
and included in FI was established with a new filling of a quartz vein. A
third fault F3 (also found in the area of Mbi) oblique to FI and F2 in the
direction N 145 sub-vertical, and joins with FI and F2, accompanied by its
vein filling. Finally a mineral lineament of direction and dip, N 120 - 35 S
(see Photo 2), has complex features that indicate it is recent.
Structurally, D1 and D2 deformations having contrary motion show
two tectonic movements, namely: tectonics of the Ebumean age (2.1 Ga)
responsible for thrusting from S-E to N-W and a second pan-African age
tectonics. We therefore suggest: a dating of metamorphic and sedimentary
rocks to coincide with the chronology of events; an elemental geochemistry
Washington Academy of Sciences
21
to trace for a Concordia diagram to have both the age of the formation and
the age of metamorphism.
On the regional level, there is a corridor of grid cut faults towards
the north of the CAR (Cornacchia and Dars, 1983) in the direction N 70°
and N 40°. This is the area of strike-slip faults of the Oubanguiides.
Other setback faults with a direction N 130° to N 160° towards the
sinistral fault affect all the structural units (Poidevin, 1991). These major
setbacks date from the pan-African phase.
As we get closer to our study area we find different faults in the
major setback of the pan-African phase described by Rolin (1992). There
are two families of faults (N 45° and N 80°) that correspond to the dextral
grid N 70° and N 40° of the pan-African phase. There are flaws running N
130° - N 140° corresponding to the sinistral transcurrent N 130° of pan-
African water. Similarly, faults running N 160° - 175° can be classified
within the family of sinistral offsets at N 160° of the pan-African phase. We
found two families of shear flaws including the first sinistral flaw (N 130°)
and the second dextral flaw (N 45°) as we have described. These two flaws
affect both formations of Mbi.
Conclusion
The CAR is a landlocked country in Central Africa. It is divided into
16 administrative prefectures, two of which are economic prefectures, and
one an autonomous commune; the prefectures are further divided into 71
sub-prefectures. Geologically CAR is a country rich in mineral resources.
Our study is located in the region of Boali, which is a town located in the
prefecture of Ombella M’poko. Boali is on the National Road 1 (RN1) about
100 km northwest of the Bangui capital of the CAR.
For this work we focused on the protocols set by previous studies of
the geology formation in the Boali region. We also considered studies of
the region by others. We include not only new information on the geologic
formation of the Boali region, but also discuss the effect that non-
exploitation of the granitoids in the area has on the region. The Boali region
has an important reserve of granitoids, which form outcroppings and
constitute the base of the region. Geologically granitoids consist essentially
of quartz and feldspar (a ferromagnetic material). In region of Boali the
Winter 2015
22
crystals form in veins which intrude into the sedimentary and metamorphic
formation. These rocks are important and useful for economic and social
development.
In the region of Boali most mining is done by artisanal gold miners.
Granitoids have never interested the people in the Boali region. We note that
in the region of Boali, none of the local residences are made with material
from the granitoids. Yet granitoids are needed for the infrastructure. In
general for the CAR and in particular for the Boali region, granitoids are
wealth ignored and abandoned. If the granitoids are exploited in the region
of Boali, then they can contribute to the buildings; for example, tiles can be
made of granitoid. Most products made with granitoids are the hardest of
materials, which offer a luxury in comparison to marble tiles.
In the petrographic plane the sedimentary rocks are quite fractured
and mingle with quartz veins, metamorphic and magmatic rocks. Magmatic
differentiation which led to the establishment of the Diorites, the
Granodiorites and Granites as well as intrusion of intermediate rocks
(Dolerite) shows a bimodal magmatism confirmed by the presence of
Granitoids and Ultrabasites.
The quartz from quartz veins can be made into glass for the
manufacture of laboratory equipment such as: burettes, beakers, and test
tubes, which are urgently needed in the Central African Republic. Quartz
veins are an asset for developing jewelry workshops. We use the quartz
from quartz veins in the manufacture of silicon pads, integrated circuits for
audio and video devices, microprocessors for computers, solar panels, and
electric watches. It is also used in gas and electronic lighters. Furthermore
it can be used in construction, for coating houses, pavements, and layering
of load-bearing seats.
Unfortunately for the CAR in general and for the region in particular
these rocks have not been exploited. We do note, however, that it is a
complex issue. Consequently the granitoids have not contributed to the
social development in the CAR and Boali. In Boali unemployed youth
might find productive work by artisanal mining of these reserves. The
government might consider implementing a policy for medium and small
business development in the areas mentioned above that might lower the
high percentage of unemployment with all its consequences, not least of
which are prostitution and violence. In our investigation in the region of
Washington Academy of Sciences
23
Boali, we conclude that currently the most important human activities are
the individual artisanal agriculture production, fishing, and hunting to
satisfy the daily needs.
References
1915. N.L. Bowen. “The later stages of the evolution of the igneous rocks” Journal of
Geology, 23, pp. 1 - 89.
1979. J.L. Poidevin. “Stratigraphic Precambrian formations of Central Africa Republic”
In: 10th symposium Geol. Afric., Montpellier, 1979-1904 / 25-27. S.L.: S.N, p. 12
(Summary).
1983. M. Comacchia, R. Dars. “A major structural feature of Africa continent. The
Central Africa Republic and Cameroon lineaments to the Gulf of Aden” Bull. Soc.
Geol. France, XIV- 1, pp. 101-109.
1985. M. Cornacchia, L. Giorgi, J.C. Lachaud. “Preliminary note on the hydrogeology of
the Bangui region. Central African Republic” 10th Congress Nat. Soc. Sav,
(Montpellier) Sciences. Fasc. VI. pp. 331-342.
1986. M. Cornacchia, L. Giorgi. “The series Precambrian sedimentary and volcano-
sedimentary Central African Republic” Ann. Mus. Roy. Afr. Central, Tervuren,
Belgium, ser. in-8°, Sci. geol., 93, p. 51.
1986. J.L. Poidevin, C. Pin. “2 GA U-Pb Zircon dating of Mbi granodiorite (Central
African Republic) and its bearing on the chronology of the Proterozoic of Central
Africa.” London: Journal of African Earth Sciences , 5, 6: pp. 581-587.
1988. J. Biandja. “Metallogenic approach of Greenstone Belt Bogoin (RCA). His gold
mineralization”. Earth Sciences. University Pierre et Marie Curie - Paris VI. 88-7. p.
345.
1989. M. Comacchia, L. Giorgi "Discrepancies and major Precambrian magmatic series
of Bogoin region. (West-Center of the Central African Republic)" , J. Afr. Earth Sci.,
vol. 9, pp. 221-226.
1991. J.L. Poidevin. “ Greenstone belts of the Central African Republic (Bandas,
Boufoyo, and Mbomou Bogoin). Contribution to the knowledge of the Precambrian
craton northern Congo ” Thesis: University Blaise Pascal Clermont-Ferrand, 458 p.
multigr.
1992. P. Rolin. “Presence of a major ductile overlap pan African age in the central part of
the Central African Republic. Preliminary Results” C.R. Acad. Sci. Paris, 315, 467-
470.
Winter 2015
24
1994. S.F. Toteu, W.R. van Schmus, J. Penaye, J.B. Nyobe. “U-Pb and Sm-Nd evidence
for Eburnian and Pan-African high-grade metamorphism in cratonic rocks of
southern Cameroon. Precambrian ’ Elsevier. Res vol. 67: pp. 321-347.
2000. J. Biandja. Private communication.
2008. G.R. Moloto-A-Kanguemba, R.I.F. Trindade, P. Monie, A. Nedelec, R. Siqueira.
“A late Neoproterozoic paleomagnetic pole for the Congo craton: Tectonic setting,
paleomagnetism and geochronology of the Nola dike swarm (Central African
Republic). Precambrian ” Elsevier. Res vol: 164: pp. 214-226.
2013. C.L. Djebebe-Ndjiguim. “Characterization of the aquifers of the Bangui urban area,
Central African Republic, as an alternative drinking water supply resource”
Hydrological Sciences Journal. 58 (8), 1760-1778.
Bios
Narcisse Bassanganam is a Master’s Student in Mineral Resource
Prospecting and Exploration at the China University of Geosciences
(Wuhan). He also received a BS in the Exploration Engineering of Mineral
and Resources from there. He has conducted field research in both China
and the Central African Republic.
Yang Mei Zhen is on the faculty of the China University of Geosciences
(Wuhan). She teaches in the Earth Resources area and directed Mr.
Bassanganam ’s work.
Prince Emilien Yedidya Danguene is an Educator Researcher at the
University of Bangui and President of the Christian Community of the
Central Africa Republic. He is the former Minister of Development of
Mining and Energy Projects.
Minfang Wang is an Associate Professor at the China University of
Geosciences (Wuhan). She has worked on many projects, including
research at the Institute of Mineralogy and Technology at Freiberg.
Washington Academy of Sciences
25
Hydrogen Clouds from Comets 266/P Christensen
and P/2008 Y2 (Gibbs) are Candidates for the Source
of the 1977 “WOW” Signal
Antonio Paris
St. Petersburg College, FL
Evan Davies
The Explorers Club, 46 East 70th St, New York, NY
Abstract
On 1977 August 15, the Ohio State University Radio Observatory
detected a strong narrowband signal northwest of the globular star cluster
M55 in the constellation Sagittarius (Sgr). The frequency of the signal,
which closely matched the hydrogen line (1420.40575 1 77 MHz), peaked
at approximately 23:16:01 EDT. Since then, several investigations into
the “Wow” signal have ruled out the source as terrestrial in origin or
other objects such as satellites, planets and asteroids. From 1977 July 27
to 1977 August 15, comets 266P/Christensen and P/2008 Y2 (Gibbs)
were transiting in the neighborhood of the Chi Sagittarii star group.
Ephemerides for both comets during this orbital period placed them at
the vicinity of the “Wow” signal. Surrounding every active comet, such
as 266P/Christensen and P/2008 Y2 (Gibbs), is a large hydrogen cloud
with a radius of several million kilometers around their nucleus. These
two comets were not detected until after 2006, therefore, the comets
and/or their hydrogen clouds were not accounted for during the “Wow”
signal emission. Because the frequency for the “Wow” signal fell close
to the hydrogen line, and the hydrogen clouds of 266P/Christensen and
P/2008 Y2 (Gibbs) were in the proximity of the right ascension and
declination values of the “Wow” signal, the comet(s) and/or their
hydrogen clouds are strong candidates for the source of the 1977 “Wow”
signal.
Introduction
On 1977 August 15 at approximately 23:16:01 EDT, the Big Ear Radio
Telescope at The Ohio State University detected an intermittent narrowband
radio signal (<10 KHz) northwest of the globular star cluster M55 in the
constellation of Sagittarius (Sgr) [ 1 ] [2] and approximately 2.5° south of the
Chi Sagittarii star group [5]. Determining the exact location where the 72-
second signal originated from in the sky was problematic because the
telescope used two separate feed horns to search for radio signals [5], The
data from the signal, moreover, were processed in such a way that it was
Winter 2015
26
difficult to establish which of the two horns detected the signal [2]. There
are, therefore, two possible right ascension values for the source of the
alleged extraterrestrial intelligence signal: 19h22m24.64s ± 10s and
19h25m17.01s± 10s and the declination was determined to be -27°03' ± 20
(Table 1) [2]. Two similar values for the signal’s frequency were assigned:
1420.356 MHz and 1420.4556 MHz. These two frequencies fall close to the
hydrogen line, which is 1420.40575177 MHz [6],
Table 1 : Right Ascension and Declination Equinox
Conversions; and Galactic Coordinates for the “Wow” Signal
(Source: Ohio State University Big Horn Report)
Declination Positive Horn Negative Horn
B1950.0 Equinox -27°03'± 20’ 19h22m24.64s ± 10s 19h25m17.01s± 10s
J2000.0 Equinox -26°57'± 20’ 19h25m31s± 10s 19h28m22s± 10s
Galactic Latitude N/A -18d53.4m± 2.1m -19d28.8m± 2.1m
Galactic Longitude N/A 1 ld39.0m± 0.91TI 1 ld54.0m± Q.9m
Previous Investigations by the Astronomical Community
Subsequent research to re-detect and identify the “Wow” signal by
The Ohio State University, the Very Large Array, and The University of
Tasmania’s Mount Pleasant Radio Observatory were null. After a search of
the area where the “Wow” signal was detected (Table 2), the Very Large
Array and The Ohio State University Radio Observatory concluded there
was strong evidence against the origin of the source as terrestrial in nature
or objects such as planets, man-made spacecraft, artificial satellites, and
radio transmissions emanating from Earth. Furthermore, the Very Large
Array proposed the intermittent “Wow” signal matched the signature of a
transiting celestial source [5], while The University of Tasmania suggested
the signal was moving with the source of the hydrogen line [7],
Anatomy of a Comet and Its Hydrogen Cloud
The distinctive parts of a comet include the nucleus, coma, dust tail,
ion tail, and a hydrogen cloud. Moderately active comets are surrounded by
a widespread cloud of neutral hydrogen atoms [4], The hydrogen is released
from the comet when ultraviolet radiation from the Sun splits water vapor
molecules released from the nucleus of the comet into the constituent
components oxygen and hydrogen [8], The size of the hydrogen cloud is
determined by the size of the comet and can extend over 100 million km in
Washington Academy of Sciences
27
width, such as the hydrogen cloud of comet Hale Bopp [9]. As a comet
approaches the Sun, its hydrogen cloud increases significantly. Since the
rate of hydrogen production from the comet’s nucleus and coma has been
calculated at 5 x 102g atoms of hydrogen every second, the hydrogen cloud
is the largest part of the comet [9]. Moreover, due to two closely spaced
energy levels in the ground state of the hydrogen atom, the neutral hydrogen
cloud enveloping the comet will release photons and emit electromagnetic
radiation at a frequency along the hydrogen line (1420.40575177 MHz)
[10].
Date of Search RA DEC
VLA 25 SEP 1995 19h21m28.1s to 19h25m48s -27°41 to -26° 18
07 MAY 1996 19h21m28.1s to 19h25m48s -27°41 to -26° 18
Ohio State U. 05 OCT 1998
09 OCT 1998
9-10 APR 1999
17-18 MAR 1999
20-21 MAR 1999
22-23 MAR 1999
Table 2: Right Ascension and Declination Observations Grid by the VLA and Ohio State
University (Source: VLA and Ohio State)
Comets 266P/Christensen and P/2008 Y2 (Gibbs)
From 1977 July 27 to 1977 August 15, Jupiter-family comets
266P/Christensen and P/2008 Y2 (Gibbs) were transiting in the vicinity of
the Chi Sagittarii star group and significantly close to the source of the
“Wow” signal (Figure 1) [3][ 1 1 ]. Of significance to this investigation, the
purported source of the “Wow” signal was fixed between the right ascension
and declination values (Table 3) of comets 266P/Christensen and P/2008
Y2 (Gibbs). On their orbital plane, moreover, 266P/Christensen was 3.8055
AU from Earth and moving at a radial velocity of +13.379 km/s; and P/2008
Y2 (Gibbs) was 4.406 AU from Earth and moving at a radial velocity of
+ 19.641 km/s (Figure 2) [3].
Winter 2015
28
Figure 1: Location of Comets 266P and P/2008 from 1977 July 27 to 1977 August 15.
(Source: The Minor Planet Center and NASA JPL Small Body Database) [11].
Table 3: Right Ascension and Declination Values for Comets P/2008 Y2
(Gibbs) and 266P/Christensen (Source: Minor Planet Center)
Washington Academy of Sciences
29
Figure 2: On 1977 August 15, comet 266P/Chrislensen was 3.8055 AU from Earth and
comet P/2008 Y2 (Gibbs) was 4.406 AU from Earth (Source: JPL Solar System Dynamics
Database) [12]
The data regarding cornets 266P/Christensen and P/2008 Y2
(Gibbs), therefore, strongly suggest either comet, or both, could be the
source of the hydrogen line signal detected by the Ohio State University on
1977 August 15. Chemicals in comets emit radio waves. The hydrogen radio
waves from a comet, such as from 266P/Christensen and P/2008 Y2
(Gibbs), travel through space akin to light. Therefore, radio telescopes,
including the Big Ear Radio Telescope at The Ohio State University, could
have intercepted them. It is noteworthy to comment, moreover, during
observations of the area by the Very Large Array and The Ohio State
University Radio Observatory (from 1995 to 1999), comet
266P/Christensen and P/2008 Y2 (Gibbs) were not in the neighborhood of
the right ascension and declination values of the “Wow” signal (Table 4)
[5], thus the hydrogen cloud from these two comets would not have been
detected. Additionally, because the period for comet 266P/Christensen is
6.63 years and P/2008 Y2 (Gibbs) is 6.8 years [3], their orbital period could
account for why the “Wow” signal was intermittent and not detected during
subsequent searches of the area.
Winter 20 15
30
Conclusions
There is noteworthy data to propose that the hydrogen signal
detected by the Big Ear Radio Telescope at The Ohio State University,
specifically 1420.356 MHz and 1420.4556 MHz, emanated from the neutral
hydrogen clouds of comets 266P/Christensen and/or P/2008 Y2 (Gibbs).
There are, conversely, many unknowns the astronomical community will
need to address to confirm the hydrogen clouds from these comets were the
source of the 1977 “Wow” signal. To date, no observations have acquired
and measured the size, mass and spectral signature, most critically, of these
two comets. Additionally, in 1977 the Big Ear Radio Telescope was
operating in drift scan mode. Consequently, if a comet (or any celestial
object) was the source of the “Wow” signal, it should have been detected in
the trailing beam after detection in the leading beam [13]. Comet
266P/Christensen will transit the neighborhood of the “Wow” signal again
on 2017 January 25 and can be located at 19h25m 15.00s and declination
-24°50' at a magnitude of +23 [3]. On 2018 January 07, comet P/2008 Y2
(Gibbs) will also transit the neighborhood of the “Wow” signal. Comet
P/2008 Y2 (Gibbs) can be located at right ascension 19h25m 17.6s and
declination -26°05' at a magnitude of +26.9 [3]. During this period, the
astronomical community will have an opportunity to direct radio telescopes
toward this phenomenon, analyze the hydrogen spectra of these two comets,
and test the authors’ hypothesis.
Washington Academy of Sciences
31
Table 4: Location of Comets 266P/Christensen and P/2008 Y2 (Gibbs) During VLA and
Ohio State University Observations (Source: The Minor Planet Center)
References
1 . Shostak, Seth (2002). “Interstellar Signal from the 70s Continues to Puzzle
Researchers”, http://archive.seti.org/epo/news/features/interstellar-signal-from-the-
70s.php accessed on 01 Oct. 2015.
2. Ehman, Jerry R. (2010). “Wow! Signal 30th Anniversary Report.” North American
Astrophysical Observatory http://www.bigear.org/Wow30th/wow30th.htm accessed
on 14 Oct. 2015.
3. The International Astronomical Union Minor Planet Center, Database: MPEC 2009-
A03 P/2008 Y2 (Gibbs); MPEC 2008-U27 266P/Christensen.
http://www.minorplanetcenter.net/ accessed on 2 1 Nov. 2015.
4. Centre for Astrophysics and Supercomputing. Cometary Hydrogen Cloud, COSMOS,
Swinburne Astronomy,
http://astronomy.swin.edu.au/cosmos/C/Cometarv+Hvdrogen+Cloud accessed on 12
Oct. 2015.
5. Gray, Robert; Marvel, Kevin (2001). “A VLA Search for the Ohio State ‘Wow’”.
ApJ, 546 (2001) pp. 1171-1177 http://www.bigear.org/Gray-Marvel.pdf accessed on
4 Nov. 2015.
6. Chaisson, Eric, and McMillan, Steve. (2005) Astronomy Today. 7th edition. Upper
Saddle River, NJ. Pearson/Prenlice Hall pp. 458-459.
Winter 2015
32
7. Gray, Robert H.; Ellingsen, Simon (2002). “A Search for Periodic Emissions at the
Wow Locale”. ApJ, 578 (2002) pp. 967-971.
8. Palen, Stacy. (2012) Understanding Our Universe, 2nd edition, New York, W.W.
Norton pp 228-230.
9. Lang, Kenneth R (2010). Hydrogen Cloud of a Comet. NASA ’s Cosmos.
https://ase.tufts.edu/cosmos/view picture.asp?id=1291 accessed on 01 Sept. 2015.
10. Term, Joe. (2015) “Hendrik C. Van De Hulst. The Bruce Medalists”. http://phys-
astro.sonoma.edu/brucemedalists/vandeHulst accessed on 01 Sept. 2015.
1 1. Comet Base Observations Catalogue for P/2008 Y2 (Gibbs); 266P/Christensen.
http://cometbase.net/en/observation/index accessed 13 Sept. 2015.
12. Jet Propulsion Laboratory Small Bodies Database. Ephemerides and Orbital
Solutions for P/2008 Y2 (Gibbs); 266P/Christensen http://ssd.ipl.nasa.gov/sbdb.cgi
accessed on 13 Sept. 2015.
13. Private Communication, (2015) Childers, Russ, Chief Observer at the OSU Radio
Observatory, 1989-1997.
Bios
Antonio Paris is a Professor of Astronomy at St. Petersburg College, FL;
the Director of Planetarium and Space Programs at the Museum of Science
and Industry in Tampa, FL; and the Chief Scientist at the Center for
Planetary Science - a science outreach program promoting astronomy,
planetary science, and astrophysics to the next generation of space
explorers. He is a member of the Washington Academy of Sciences, the
American Astronomical Society, the St. Petersburg Astronomy Club, FL;
and the author of two books, Aerial Phenomena and Space Science.
Evan Davies is a fellow of both the Royal Geographical Society and The
Explorers Club, and his popular space science writing has appeared in
Wiley publications as well as Archaeology and Spaceflight magazines. He
is the author of Emigrating Beyond Earth: Human Adaptation and Space
Colonization and has held a lifelong interest in space exploration.
Washington Academy of Sciences
33
Affine Geometry, Planck Length and Cosmic
Acceleration
George L. Murphy
Tallmadge, OH
Abstract
In Ihe 1940s Schrodinger developed a generalization of Einstein’s metric
gravitational theory based on a purely affine geometry. Today there are some
reasons to give this theory renewed attention. First, it is another step along the
path that Einstein pioneered in abandoning a priori assumptions about the
geometry of the world. Second, Schrodinger’ s theory offers the prospect of
dealing with the breakdown of the metric concept at the Planck scale while
retaining the continuum. And third, the requirement that the cosmological
constant cannot vanish in this theory means that the cosmic acceleration which
has recently been discovered can be included in a natural way with this
approach, and that the problem of a large vacuum energy can be resolved.
Introduction
A scientific theory that does not predict novel phenomena or correlate
known facts better than its competitors will be relegated to the history of
science museum. It may, however, return to active duty if it helps to explain
new data. Today Schrodinger’ s affine field theory from the 1940s deserves
such reconsideration. 1 It may help to explain cosmic acceleration and
provide a basis for quantized gravitation as a result of an advance beyond
metric-based general relativity.
This theory received inadequate attention, or the wrong type of
attention, when it was proposed. Many physicists considered it to be only a
variant of a theory Einstein was then developing in which a non-symmetric
part was added to the metric tensor.2 Today we can see that a theory in which
metric is a secondary concept can explore territory that is closed to a theory
in which it is fundamental.
When Schrodinger proposed his theory many relativists viewed the
cosmological term negatively. Pauli, for example, rejected it because it
required a non-vanishing cosmological constant. 3 Now we know that
cosmic expansion is accelerating in a way that is compatible with a
cosmological term in the gravitational field equations.4
W inter 20 1 5
34
Attempts to extend general relativity like those of Schrodinger and
Einstein were also burdened by expectations that they could be unified field
theories encompassing all physical phenomena. Einstein’s hope that a
successful theory of this type would eliminate what he saw as unattractive
features of quantum mechanics gave many physicists the impression that
the whole line of work was essentially reactionary.5
Here we eschew any expectation that an affine theory can, by itself,
provide a unified explanation of physical interactions. While it does
generalize Einstein’s theory of gravitation, its main interest is that it
provides a broader geometric framework for further work. We will begin by
considering reasons for generalizing the geometry that Einstein used in his
1915 theory, and then explore the possibilities connected with quantum
gravity and dark energy.
Is Riemannian Geometry Necessary?
The name “geometry”, from Greek words for “earth” and
“measure”, shows the discipline’s origin in practical concerns. But
geometry also became the theoretical system of Euclid. It was long assumed
that this system described the world correctly, and Kant’s view that our
minds must perceive the world that way put a sophisticated seal on the idea.6
But failures to prove Euclid’s parallel postulate finally moved
mathematicians to realize that it could be replaced by another assumption,
and thus to develop non-Euclidean geometries.7 Further progress resulted in
the Riemannian differential geometry that Einstein used in his gravitational
theory. He did not impose a global geometry a priori but made the local
character of space-time something to be determined by physical
measurements. Geometry and physics were united, an achievement that
Weyl symbolized in the equation “Pythagoras + Newton = Einstein.”8
General relativity uses a metric geometry: The local properties of
space-time are completely specified by the metric tensor gov. It is, however,
possible to consider more general differential geometries, as Weyl did
within three years of the introduction of Einstein’s theory in an attempt to
include electromagnetism.9 Other generalizations then followed.10
Attempts to extend Einstein’s unification of physics and geometry
are viewed most helpfully in the spirit of Klein’s Erlanger Programme
Here different geometries are considered in terms of the transformation
Washington Academy of Sciences
35
groups that they allow. We can begin with a topological space whose
meaningful properties are invariant under all continuous transformations,
and then specialize by limiting this group, allowing new properties to
emerge. The concepts of lines and points as their intersections are invariant
under projective transformations and are therefore meaningful in projective
geometry. A structure of parallelism can be added to yield affine geometry,
and then metric properties can be introduced.
By assuming a metric geometry of space-time we add concepts and
specialize. We need not use a geometry more complex than necessary for
physics, but insistence that the geometry of the world must be Riemannian
is in the same spirit as the old idea that the geometry of the world must be
Euclid’s.
These arguments will appeal to those who believe that Einstein’s
geometric view of gravitation was fundamental. Physicists who see it as just
one way of interpreting his field equations will question the value of, for
example, allowing torsion, a skew-symmetric part of the affine connection.
This seems to be, as Weinberg argues, “just a tensor,” just one more field.12
But in generalizing the geometry of Einstein’s gravitational theory we are
not adding torsion but removing conditions on the geometry that imply the
vanishing of torsion.
The fundamental question for physics, however, is whether a
geometric approach will facilitate our understanding of our observations of
the world. When we consider that possibility we will see some of the
concrete advantages offered by an affine theory.
Schrodinger’s Affine Theory
The basic assumption of this theory is a four-dimensional manifold
with an affine connection T^v which is not assumed to be symmetric and
which gives the change in a vector A tt when displaced by a coordinate
increment dxv: 8 A A = -r“vAA dxv .
In the usual way, the covariant derivative of a vector field A A can be
defined as Af'a = AMa + T^aAA and a curvature tensor can be constructed as
shown by Equation (1):
Winter 2015
36
dP - pA _ p/2 i p// p« _ p// p«
ovp 1 o\\p 1 opy ' 1 apv ov avl op'
(i)
This has two independent contractions. The first, given by Equation (2),
Ra* = Kp
r7'
CJV,fl
rM
op,v
+ va —
' 1 apl o\’ A av 1 op '■
(2)
is a generalization of the Ricci tensor of Riemannian geometry. The second,
given by Equation (3):
o = bu =rw -r;/ (3)
pov poy pv,o"> V '
which vanishes in a Riemannian space, is completely antisymmetric.
Metric concepts have not yet been introduced and there is no way to
compare lengths along different curves or to measure angles. We know,
however, that at some scales the concepts of length and time are meaningful
in our universe. In order to represent them we need to have a metric tensor,
a symmetric second rank tensor gm/ , to define a magnitude ds associated
with a displacement dx A via the generalization of the Pythagorean theorem
ds 2 = gfJvdxfJdxv .
If g is not to be a foreign body within the theory then we must use a
symmetric second rank tensor that has already been defined, and the only
possibility at this stage is the symmetric part of Rm,. Thus gm/ must be
proportional to R(av). We can write this relationship more suggestively as
Equation (4):
^«n.) = Agov W
with A a constant. If ds is to have the same dimension as dx , dimensional
analysis with powers of dxA shows that R(av) has dimension -2 so that A
must also have dimension -2. In metric language it must have the
dimensions of an inverse length squared.
The similarity between Eq. (4) and Einstein’s vacuum equations
with a cosmological term can hardly be missed. (The symbol A was not
chosen at random.) To see the significance of this more fully, we need to
look briefly at Schrodinger’s formulation of what he called “the final affine
field laws.”
Washington Academy of Sciences
37
To obtain dynamical equations for the field variables T^v we use
Hamilton’s principle with an appropriate Lagrange density. At this point we
have no way to raise and lower indices, so the simplest density can be
written as Eq. (5):
L = (2 / A)
(5)
where A is a constant which will give the action the proper dimensions. (It
is also possible to use R + (l/4)£? in Eq. (5) to achieve projective
invariance.)13
Following Schrodinger, we define the tensor density g"1 by
9L / 9R„„ = g"1'
(6)
and form the covariant and contravariant tensors g and gftv associated
with it. The Euler-Lagrange equations of the variational principle then give
Eq. (7):
g^-*r^g«v-*r^,9^ = °. (7)
where *r" , Schrodinger’s “star affinity”, is an abbreviation for
C + (2 / 3)S“Yp and Y „ is (1/2)
PP
torsion.
T^a T^a
pa ap
a contraction of the
The defining Eq. (5) is equivalent to Eq. (8)
so that Eq. (6) can be written as an equation involving only the affinity and
the contracted curvature tensor. In the limiting case in which is
symmetric, Eq. (6) is the equation satisfied by the metric tensor in
Riemannian geometry. It can be solved to give the as functions of g
and its derivatives, the Christoffel brackets. R is then the symmetric Ricci
tensor of general relativity, and Eq. (4) and Eq. (7) are seen to be identical,
the vacuum Einstein equations with a cosmological term.
Winter 2015
38
The Demise of Metric
The metric concept is meaningful in our universe only at sufficiently
large scales. The uncertainty principle of quantum theory and the effect of
gravitation on clock rates, together with limited resolution of a clock due to
its size, show that below certain limits time intervals cannot be measured.14
To measure a time interval we must use a clock. The time-energy
uncertainty relation tells us that the time t taken for the measurement and
the uncertainty in the clock’s energy must satisfy tAE > fi . On the other
hand, the gravitational field of the clock’s energy of a distance R will result
in a change in a time interval t given by St / 1 ~ GE / c4R , where E includes
both the original rest energy of the clock and AE. t can be no larger than R/c
if the parts of the clock are to communicate with one another. When these
results are combined we find that St > GJi / c5t . Since the measurement is
of no value unless St <t , t must be greater than T*
TiG/c 5
~ 1 0”43 s
the Planck time. The corresponding limit for spatial dimensions, the Planck
length, is f * = cf* ~ 1 0-35 m .
Planck noted when he introduced the constant h that with c and G it
implied natural units of length, time and mass,15 but the significance of these
units seems not to have been given much consideration in the period when
Schrodinger was developing his theory. As the previous paragraph shows,
T* and L* are not just results of dimensional analysis but basic limits on
measurement of space-time intervals. Metric concepts are valid only for
sufficiently large regions, a possibility already foreseen by Riemann.16
It has been suggested by some authors that various types of discrete
structure for space-time should be considered at small scales.17 But there are
problems with abandoning the continuum concept and giving space-time an
atomistic character. The full group of coordinate transformations of the
continuum cannot be adequately approximated in a discrete model of space-
time.18 It seems wise to take a more conservative approach and consider
space-time as a continuous manifold in which the metric concept breaks
down at sufficiently small scales.
One way to understand the failure of space-time’s metrical character
in the context of affine theory begins by noting that a metric tensor defined
by Eq. (4) depends on the first derivatives of the connection. (This reverses
Washington Academy of Sciences
39
the relationship of Riemannian geometry, in which the Christoffel brackets
involve derivatives of the metric.) If T^v were continuous but not
differentiable in a region then gm. would not be defined there. The affine
connection in those regions could not obey differential equations and would
have a fractal structure. 19 One consequence of this would be scale
invariance: No constant with the dimensions of length would be involved.
It is not difficult to find functions that are everywhere continuous
but differentiable nowhere, or only on a set of points of measure zero, in a
domain: Weierstrass first published an example of such a function, the
Fourier series / (/) = ^Jb" sin ycin7rt^ , where n ranges from 1 to go and ab >
1 + 3ti/2.20 If the connection coefficients were represented by such functions
then their behavior in their function space would be somewhat like
Brownian motion or turbulence.
Suppose that this were the case for sufficiently small regions of
space-time. (“Small” can only be defined from the outside, since there is
no metric inside these regions.) The metric concept would break down if
attempts were made to explore such regions, and we have already seen that
that is actually the case for time and length scales below the Planck values.
On the other hand, scale invariance would be broken by the presence of the
length |A| ' “ . We will consider question of the spatio-temporal scale at
which the metric concept fails when we discuss cosmology in the next
section.
Palmer has proposed a new approach to quantum theory based on
the ideas “that states of physical reality belong to, and are governed by, a
non-computable fractal subset I of state space, invariant under the action of
some subordinate deterministic causal dynamics D” and that “gravity plays
a key role in generating the fractal geometry off”21 An affine theory seems
to have the potential to provide such a geometry. But whether that proposal
is pursued, it seems clear that an affine approach can ensure that one of the
basic requirements for an adequate theory of quantum gravity is satisfied.
Winter 2015
40
The Necessity of a Cosmological Term
The way in which we have led up to Eq. (8) brings out the inevitable
appearance of the cosmological term. It is instructive to proceed this way
because that term has been the object of controversy. Einstein added this
term to his original equations to make a static universe possible and then
dropped it when it was found that the universe is expanding. Between
Hubble’s discovery of cosmic expansion in 1929 and the realization in the
late 1990s that this expansion was indeed speeding up, many workers in
general relativity and cosmology dismissed the cosmological term.
A number of theoretical models of dark energy have been proposed
to account for the acceleration of cosmic expansion, but present data are
compatible with Einstein's 1917 cosmological term.22 The simplest solution
of Eq. (8) is the well-known de Sitter metric which can be written as Eq.
(9):
ds2 =-dt2 + exp(2Ht)dl2 , (9)
where c = 1 now (so that length and time scales are identical), dl is the
Euclidean spatial line element and H = (A/3) “ is the Hubble constant.
The fact that the purely affine theory requires a cosmological term must
count in its favor as prediction of a “novel fact,” one that was not assumed
in the formulation of the theory. Most other dark energy models have been
introduced precisely to explain cosmic acceleration and cannot therefore
count it as a prediction.
Quantum theory, however, presents the problem of a cosmological
constant that is far too large. The cosmological term in the gravitational
field equations has the form of a stress-energy tensor for a fluid whose
pressure is negative and equal in magnitude to its energy density, and the
vacuum energy of quantum fields has just this form. This effective
cosmological constant obtained from quantum field theory cannot be
reconciled with observations. Zel’dovich’s23 calculation of the vacuum
stress-energy tensor for an assortment of boson and fermion fields with a
cutoff on the order of the proton’s Compton wavelength gave a value 44
orders of magnitude larger than what observations at that time would allow.
A more fundamental cutoff is defined by the Planck length.
Washington Academy of Sciences
41
The resulting vacuum energy in Einstein’s equations gives a model
universe that, according to Eq. (9), expands exponentially with a
characteristic time on the order of T* ~ 10 43 s. Of course this is
unacceptable.
But things could be different when we consider this vacuum energy
in the context of affine theory. The vacuum energy density p gives rise to
an effective cosmological constant 87 ip, and if we were combine this with
A to give an effective cosmological constant we get Eq. (10):
We could then choose the value of A to “renormalize” the effective
cosmological constant to a value A' in accord with observations, as
Shifflett has suggested.24 Since p has a large positive value, A would have
to have a negative value of nearly the same magnitude.
Choosing such a value for A is not completely arbitrary. The
cosmological constant defines a fundamental length and time |A| ' ~ which
would break the scale invariance of the fractal regime discussed in the
previous section. If that is close to the cutoff for calculation of vacuum
energy then A and A' would be of the same order of magnitude.
Affine theory has a universal standard of length, |A| 1 “ . In the 1930s
Eddington gave this as a reason to retain “the cosmical constant” even after
the original motive for it had disappeared.25 It seemed obvious then that this
length would be of cosmological size. The idea that there are two
fundamental lengths, one provided by A and the other defined by the
fundamental constants h,c, and G, might have raised suspicions if the
physical significance of the Planck length had been given more attention at
that time. We can see now the possibility that the two fundamental lengths
are approximately the same.
Winter 20 15
42
Prospects for Further Progress
Eqs. (7) and (8) are the field equations of Schrodinger’s theory,
which reduce to Einstein’s vacuum equations with a cosmological term in
the limiting case of a symmetric connection. Since we are not pursuing a
unified field theory we can describe other fields and particles by introducing
non-geometric variables <f>A and a Lagrange density Lm which will depend
on the ®A, their derivatives and g which, in turn, is a function of T^v and
its derivatives via Eq. (8).26 If Lm is added to Eq. (5) and the procedures of
Hamilton’s principle are earned through then the full gravitational field
equations with an energy-momentum tensor defined in the standard way can
be obtained if that tensor is small in comparison with the cosmological term.
The energy densities of baryonic and dark matter are much smaller
than the magnitude of the A that we have hypothesized to renormalize
quantum vacuum energy, so this approximation makes sense for those
forms of matter. However, the vacuum energy itself is comparable in
magnitude with A. So while this approximation method has some value, it
does not enable us to shed any light on features of quantum field theory such
as ultraviolet divergences.
We have taken advantage of the possibility of a connection that is
not differentiable to suggest that the metric concept might break down
below some space-time scale, and suggest that this could be correlated with
the implication of the uncertainty principle and the gravitational effect on
clock rates that intervals below the Planck scale cannot be measured.
However, this would also mean that the Ricci tensor, which is used to form
the Lagrangian, involves derivatives of the connection and would not be
defined. The classical Hamilton’s principle could no longer be used to
derive field equations. This is not surprising if the metric concept indeed
fails below the Planck scale.
One way to proceed would be to look for an algebraic expression
which approximates the classical action Eq. (5) on scales for which the latter
is meaningful. We could then use this action in Feynman’s “sum over
histories” approach to quantum theory in order to explore the implications
of the affine theory further.
Washington Academy of Sciences
43
For now we have barely a hint of ways in which an adequate
quantum theory of matter and non-gravitational phenomena might be
developed on the basis of affine geometry. This approach does, however,
seem capable of showing in a natural way how the metric concept may fail
below the Planck scale without abandoning the continuum concept. In
addition, the requirement that the cosmological constant not vanish and the
possibility that the theory can deal with the problem of a huge quantum
vacuum energy means that the observed cosmic acceleration can be
explained in an unforced manner. The long-dormant theory that
Schrodinger proposed in the 1940s seems today to have some promise.
References
1 E. Schrodinger, Proc. R. Irish Acad. 51 A, 163 (1947); 51 A, 205 (1948); 52 A, 1(1948).
These are summarized in E. Schrodinger, Space-Time Structure, Cambridge, (1963),
Chapter XII.
2 A. Einstein, The Meaning of Relativity, 5th ed., Princeton, (1956), Appendix II, is the
final version of Einstein’s mixed affine-metric theory.
3 W. Pauli, Theoiy of Relativity, Pergamon (1958), p.225. The dismissive footnote in
L.D. Landau and E.M Lifshitz, The Classical Theoiy of Fields, revised 2d ed.,
Addison- Wesley, (1962) may also be noted.
4 S.W. Allen et al., Mon. Not. R. Astron. Soc. 383, 879 (2008); Vikhlinin, A. et a!.:
Astrophys. J. 692, 1060 (2009).
5 Schrodinger’s attitude toward this issue was more nuanced than Einstein’s. See, e.g.,
his comment on p. 160 of Space-Time Structure.
6 P. Guyer, Kant and the Claims of Knowledge, Cambridge (1987), pp. 359-360.
7 C.B. Boyer and U.C. Merzbach, A Histoiy of Mathematics, 2d ed., John Wiley, New
York (1989), pp.580-583.
8 H. Weyl, Space-Time-Matter, 4th ed., Dover, (1950), p.228. Weyl represented Einstein
as a bracketing of Pythagoras and Newton but I have written this as an equation to
simplify typography.
9 Ibid., pp.282-312.
10 A.S. Eddington, The Mathematical Theoiy of Relativity, 2d ed., Cambridge, (1924).
Chapter VII and Supplementary Notes 13 and 14 provide a survey of some of this
work.
1 1 Boyer and Merzbach, A History of Mathematics, pp.6 11-613.
Winter 2015
44
12 S. Weinberg, Physics Today 59.4, 16 (2006). For Weinberg’s more extended
statement on the geometric approach to general relativity see his Gravitation and
Cosmology >, John Wiley (1972), pp.vi - viii and 147.
13 G.L. Murphy, Phys. Rev. Dll, 2752(1975).
14 G.L. Murphy, Am. J. Phys. 42, 958(1974).
13 M. Planck, Sitzungsber. Dent. Akad. Wiss. Berlin. Kl. Math. - Phys. 440-480 (1899).
16 B. Riemann, Abhandhmgen d. K. Gesells. zu Gottingen 13, 133 (1854).
17 R. Penrose, The Road to Reality: A Guide to the Laws of the Universe, Alfred A.
Knopf, (2005), pp. 958-962. The negative result of one search for such structure is
reported at http://arxiv.org/abs/1512.01216 .
Is B.S. DeWitt, ’’The Quantization of Geometry”, In L. Witten, (ed.) Gravitation: An
Introduction to Current Research, John Wiley, (1962), pp. 372-373.
19 B.J. West, M. Bologna, P. Grigolini, Physics of Fractal Operators, Springer, (2003),
Chapter 1.
20 K. Weierstrass, Abhandhmgen aus der Functionenlehre, Springer, (1886), p.97;
http://planetmath.org/encyclopedia/WeierstrassFunction.html .
21 T.N. Palmer, Proc. Roy. Soc. A465, 3165 (2110).
22 Cf. endnote 4.
23 Y.B. Zel’dovich, Sov. Phys. Usp. 11, 381 (1968).
24 J.A. Shifflett Gen. Rel. & Grav. 40, 1745 (2008); 41, 1865 (2009).
25 A. Eddington, The Expanding Universe, Macmillan (1933), p. 148.
26 G.L. Murphy, “An Affine Approach to General Relativity”, presented at the Third
Australasian Conference on General Relativity, University of Adelaide, Adelaide,
South Australia, 30 January 1976.
Bio
George L. Murphy received a Ph.D. in physics from Johns Hopkins
University in 1972 and has published physics papers primarily on aspects
of general relativity and cosmology. After teaching in colleges for 12 years
he began theological study, receiving a M.Div. from Wartburg Seminary,
and served as a parish pastor until retirement. He has written extensively on
science-theology relationships.
Washington Academy of Sciences
45
Membership List for 2015
M : Member, LM — Life Member, F = Fellow, LF = Life Fellow, EM = Emeritus
Member, EF = Emeritus Fellow
ABEL, DAVID (Dr.) 1 13 Hedgewood Drive, Greenbelt MD 20770-1610 (LM)
ADKINS, MICHAEL K (Mr.) 4143, Elizabeth Lane, Annandale VA 22003 (M)
ANTMAN, STUART (Dr.) University of Maryland, 2309 Mathematics Building,
College Park MD 20742-4015 (EF)
APPETITI, EMANUELA PO Box 25499, Washington DC 20027 (LM)
APPLE, DA1NA DRAVNIEKS National Capital Society of American Foresters, PO
Box 9288, Arlington VA 22219 (M)
ARLE, KATHY (Ms.) 3049 Heather Lane, Falls Church VA 22044 (M)
ARSEM, COLLINS (Mr.) 3144 Gracefield Rd Apt 117, Silver Spring MD 20904-
5878 (EM)
ARVESON, PAUL T. (Mr.) 6902 Breezewood Terrace, Rockville MD 20852-4324
(F)
BARBOUR, LARRY L. (Mr.) Pequest Valley Farm, 585 Townsbury Road, Great
Meadows NJ 07838 (M)
BARWICK, W. ALLEN (Dr.) 13620 Maidstone Lane, Potomac MD 20854-1008
(F)
BEAM, WALTER R. (Dr.) 4804 Wellington Farms Drive, Chester VA 23831 (F)
BECKER, EDWIN D. (Dr.) Bldg. 5, Rm. 128, Nat. Institutes Of Health, Bethesda
MD 20892-0520 (EF)
BEMENT, ARDEN (Dr.) National Science Foundation, 4201 Wilson Boulevard,
Arlington VA 22230 (EF)
BERLEANT, DANIEL (Dr.) 12473 Rivercrest Dr., Little Rock AR 72212 (M)
BERRY, JESSE F. (Mr.) 2601 Oakenshield Drive, Rockville MD 20854 (M)
BIGLARI, HAIK (Dr.) Sr. Director of Engineering, Fairchild Controls, 540 Highland
Street, Frederick MD 21701-7672 (M)
BIONDO, SAMUEL J. (Dr.) 10144 Nightingale St., Gaithersburg MD 20882 (EF)
BODSON, DENNIS (Dr.) 233 N. Columbus Street, Arlington VA 22203 (F)
BOISVERT, RONALD F. (Dr.) Mail Stop 8910, Nat. Institute of Standards and
Technology, 100 Bureau Drive, Gaithersburg MD 20899-8910 (F)
BOYCE, GEORGE (Mr.) 7 Greenway PL, Greenbelt MD 20070 (M)
BRISKMAN, ROBERT D. (Mr.) 61 Valerian Court, North Bethesda MD 20852 (F)
BUFORD, MARILYN (Dr.) 3073 White Birch Court, Fairfax VA 22031 (F)
CASTILLO, YOLANDA F. (Ms.) 1717 Pennsylvania Ave NW Ste 1025,
Washington DC 20008-3951 (M)
CAWS, PETER J (Dr.) Apt. 230, 2475 Virginia Ave. N.W., Washington DC 20037
(M)
CHRISTMAN, GERARD (Mr.) 6109 Berlee Drive, Alexandria VA 22312 (F)
CIORNEIU, BORIS (Dr.) 20069 Great Falls Forest Dr., Great Falls VA 22066 (M)
CIUCA, LIVIU BOGDEN (Mr.) Aleea Scolii, no.2, Galati, Bucharest, Romania (M)
CLINE, THOMAS LYTTON (Dr.) 13708 Sherwood Forest Drive, Silver Spring MD
20904 (F)
Winter 2015
46
COFFEY, TIMOTHY P. (Dr.) 976 Spencer Rd., McLean VA 22102 (F)
COHEN, MICHAEL P. (Dr.) 1615 Q. St. NW T-l, Washington DC 20009-6310
(LF)
COLE, JAMES H. (Mr.) 9709 Katie Leigh Ct, Great Falls VA 22066-3800 (F)
CURRIE, S.J., CHARLES L. (Rev.) Jesuit Community, Georgetown University,
Washington DC 20057 (EF)
DANCKWERTH, DANIEL 419 Beach Drive, Annapolis MD 21403-3906 (M)
DANNER, DAVID L. (Dr.) 1364, Suite 101, Beverly Road, McLean VA 22101 (F)
DAVIS, ROBERT E. (Dr.) 1793 Rochester Street, Crofton MD 21114 (F)
DEAN, DONNA (Dr.) 367 Mound Builder Loop, Hedgesville WV 25427-7211 (EF)
DEDRICK, ROBERT L. (Dr.) 21 Green Pond Rd, Saranac Lake NY 12983 (EF)
DIMITOGLOU, GEORGE (Dr.) 11053 Seven Hill Lane, Potomac MD 20854 (M)
DONALDSON, JOHANNA B. (Mrs.) 3020 North Edison Street, Arlington VA
22207 (EF)
DURRANI, SAJJAD (Dr.) 17513 Lafayette Dr, Olney MD 20832 (EF)
EDINGER, STANLEY EVAN (Dr.) Apt #1016, 5801 Nicholson Lane, North
Bethesda MD 20852 (EM)
EPHRATH, ARYE R. (Dr.) 5467 Ashleigh Rd., Fairfax VA 22030 (M)
ERICKSON, TERRELL A. (Ms.) 4806 Cherokee St., College Park MD 20740-1865
(M)
ETTER, PAUL C. (Mr.) 16609 Bethayres Road, Rockville MD 20855 (F)
FASANELLI, FLORENCE (Dr.) 471 1 Davenport Street, Washington DC 20016
(EF)
FAULKNER, JOSEPH A. (Mr.) 2 Bay Drive, Lewes DE 19958 (F)
FRANKLIN, JUDE E. (Dr.) 7616 Carteret Road, Bethesda MD 20817-2021 (F)
FRASER, GERALD (Dr.) 581 1 Cromwell Drive, Bethesda MD 20816 (M)
FREEMAN, ERNEST R. (Mr.) 5357 Strathmore Avenue, Kensington MD 20895-
1160 (ELF)
GAGE, DOUGLAS W. (Dr.) XPM Technologies, 1020 N. Quincy Street, Arlington
VA 22201-4637 (M)
GAITHER, VERLESE P. (Ms.) 10301 Musket Court, Fort Washington MD 20744
(M)
GAUNAURD, GUILLERMO C. (Dr.) 4807 Macon Road, Rockville MD 20852-
2348 (EF)
GIBBON, JOROME (Mr.) 3 1 1 Pennsylvania Avenue, Falls Church VA 22046 (F)
GIBBONS, JOHN H. (Dr.) Resource Strategies, P.O. Box 379, The Plains VA 20198
(EF)
GIFFORD, PROSSER (Dr.) 59 Penzance Rd, Woods Hole MA 02543-1043 (F)
GIORDANO, JAMES (Dr.) Neuroethics Studies Program, Pellegrino Center for
Clinical Bioethics, Georgetown University Med Center, Washington DC 20057
(M)
GLUCKMAN, ALBERT G. (Mr.) 18123 Homeland Drive, Olney MD 20832-1792
(EF)
GOLDSTON, DAVID (Mr.) 816 N. Highland St., Arlington VA 22201 (M)
Washington Academy of Sciences
47
GRAY, MARY (Prof.) Department of Mathematics, Statistics, and Computer
Science, American University, 4400 Massachusetts Avenue NW, Washington DC
20016-8050 (F)
GRIER, REBECCA (Dr.) 6300 Stevenson Ave #501, Alexandria VA 22304 (M)
HACK, HARVEY (Dr.) 176, Via Dante, Arnold MD 21012-1315 (F)
HAIG, SJ, FRANK R. (Rev.) Loyola University Maryland, 4501 North Charles St,
Baltimore MD 21210-2699 (F)
HARR, JAMES W. (Mr.) 180 Strawberry Lane, Centreville MD 21617 (EF)
HAZAN, PAUL 14528 Chesterfield Rd, Rockville MD 20853 (F)
HEANEY, JAMES B. 6 Olive Ct, Greenbelt MD 20770 (M)
HENDEE, JAMES (Dr.) 51 1 SE 13th Court, Pompano Beach FL 33060 (M)
HERBST, ROBERT L. (Mr.) 4109 Wynnwood Drive, Annadale VA 22003 (LF)
H1BBS, EUTHYMIA D. (Dr.) 7302 Durbin Terrace, Bethesda MD 20817 (M)
HIETALA, RONALD (Dr.) 6351 Waterway Drive, Falls Church VA 22044-1322
(M)
HILL II, RICHARD E. (Mr.) 4360 Lee Hwy #204, Arlington VA 22207 (M)
HOFFELD, J. TERRELL (Dr.) 1 1307 Ashley Drive, Rockville MD 20852-2403 (F)
HOLLAND, PH.D., MARK A. (De.) 201 Oakdale Rd., Salisbury MD 21801 (M)
HOLLINSHEAD, ARIEL (Dr.) 23465 Harbor View Rd. #622, Punta Gorda FL
33980-2162 (EF)
HONIG, JOHN G. (Dr.) 7701 Glenmore Spring Way, Bethesda MD 20817 (LF)
HORLICK, JEFFREY (Mr.) 8 Duvall Lane, Gaithersburg MD 20877-1838 (F)
HOROWITZ, EMANUEL (Dr.) Apt 618, 3100 N. Leisure World Blvd, Silver Spring
MD 20906 (EF)
HOROWITZ, SHARYN (Ms.) 217 Katie Court, Falls Church VA 22046 (M)
HOWARD, SETH ANNE (Dr.) 5526 Green Dory Lane, Columbia MD 21044 (LF)
HOWARD-PEEBLES, PATRICIA (Dr.) 7501 Virginia Parkway 2312, McKinney
TX 75071 (EF)
HURDLE, BURTON G. (Dr.) 3440 South Jefferson St., Apt. 356, Falls Church VA
22041 (F)
IKOSSI, KIKI (Dr.) 6275 Gentle LN, Alexandria VA 22310 (F)
JACOX, MARILYN E. (Dr.) 16665 Concord St Ste 206, Kenington MD 20895-2525
(F)
JARRELL, H. JUDITH (Dr.) 9617 Alta Vista Ter., Bethesda MD 20814 (F)
JAYARAO, ARUNDHATI (Dr.) 8811 Trafalgar Ct, Springfield V A 22151 (M)
JOHNSON, EDGAR M. (Dr.) 1384 Mission San Carlos Drive, Amelia Island FL
32034 (LF)
JONG, SHUNG-CHANG (Dr.) 8892 Whitechurch Ct, Bristow VA 20136 (LF)
JORDANA, ROMAN DE VICENTE (Dr.) Batalla De Garellano, 15, Aravaca,
28023, Madrid, Spain (EF)
KADTKE, JAMES (Dr.) Apt. 824, 1701 16th St. NW, Washington DC 20009-3131
(M)
KAHN, ROBERT E. (Dr.) 909 Lynton Place, Mclean VA 22102 (F)
KAPETANAKOS, C. A. (Dr.) 443 1 MacArthur Blvd, Washington DC 20007 (EF)
KATZ, ROBERT (Dr.) 3310 N. Leisure Blvd #530, Silver Spring MD 20906 (EF)
Winter 2015
48
KAUFHOLD, JOHN (Dr.) Suite 1200, 4601 N. Fairfax Dr, Arlington VA 22203
(M)
KAY, PEG (Ms.) 6111 Wooten Drive, Falls Church VA 22044 (LF)
KEEFER, LARRY (Dr.) 7016 River Road, Bethesda MD 20817 (F)
KE1SER, BERNHARD E. (Dr.) 2046 Carrhill Road, Vienna VA 22181-2917 (LF)
KLINGSBERG, CYRUS (Dr.) Apt. LI 84, 500 E. Marylyn Ave, State College PA
16801-6225 (EF)
KLOPFENSTEIN, REX C. (Mr.) 4224 Worcester Dr., Fairfax VA 22032-1 140 (LF)
KRUEGER, GERALD P. (Dr.) Krueger Ergonomics Consultants, 4105 Komes
Court, Alexandria VA 22306-1252 (EM)
KUO, CHUN-HUNG (Dr.) 4637 Knight Place, Alexandria VA 223 1 1 (M)
LAWSON, ROGER H. (Dr.) 10613 Steamboat Landing, Columbia MD 21044 (EF)
LEDGER, SAM (Mr.) 420 7th Street NW, Apt 903, Washington DC 20004 (M)
LE1BOWITZ, LAWRENCE M. (Dr.) 3903 Laro Court, Fairfax VA 2203 1 (LF)
LEMKIN, PETER (Dr.) 148 Keeneland Circle, North Potomac MD 20878 (EM)
LESHUK, RICHARD (Mr.) 9004 Paddock Lane, Potomac MD 20854 (M)
LEWIS, DAVID C. (Dr.) 27 Bolling Circle, Palmyra VA 22963 (F)
LEWIS, E. NEIL (Dr.) Malvern Instruments, Suite 300, 7221 Lee Deforest Dr,
Columbia MD 21046 (M)
LIBELO, LOUIS F. (Dr.) 9413 Bulls Run Parkway, Bethesda MD 20817 (LF)
LONDON, MARILYN (Ms.) 3520 Nimitz Rd, Kensington MD 20895 (F)
LONGSTRETH, III, WALLACE I (Mr.) 8709 Humming Bird Court, Laurel MD
207231254 (M)
LOOMIS, TOM H. W. (Mr.) 11502 Allview Dr., Beltsville MD 20705 (EM)
LUTZ, ROBERT J. (Dr.) 6031 Willow Glen Dr, Wilminton NC 28412 (EF)
LYON, HARRY B. (Mr.) 7722 Northdown Road, Alexandria VA 22308-1329 (M)
LYONS, JOHN W. (Dr.) 7430 Woodville Road, Mt. Airy MD 21771 (EF)
MALCOM, SHIRLEY M. (Dr.) 12901 Wexford Park, Clarksville MD 21029-1401
(F)
MANDERSCHEID, RONALD W. (Dr.) 10837 Admirals Way, Potomac MD 20854-
1232 (LF)
MASON, LANCE (Dr.) 1212 Calla Cerrito, Santa Barbara CA 93101 (M)
MCFADDEN, GEOFFREY B (Dr.) 20117 Darlington Drive, Montgomery Village
MD 20886 (M)
MENZER, ROBERT E. (Dr.) 90 Highpoint Dr, Gulf Breeze FL 32561-4014 (EF)
MESSINA, CARLA G. (Mrs.) 9800 Marquette Drive, Bethesda MD 20817 (F)
METAILIE, GEORGES C. (Dr.) 18 Rue Liancourt, 75014 Paris, FRANCE (F)
MIELCZAREK, EUGENIE A. (Dr.) 3181 Readsborough Ct, Fairfax VA 2203 1 -
2625 (F)
MILLER, JAY H. (Mr.) 8924 Ridge Place, Bethesda MD 20817-3364 (M)
MILLER II, ROBERT D. (Dr.) The Catholic University of America, 10918 Dresden
Drive, Beltsville MD 20705 (M)
MORGOUNOV, ALEXEY (Dr.) CIMMYT, P.K. 39, Emek, Ankara 0651 1, Turkey
(M)
MORRIS, JOSEPH (Mr.) PO Box 3005, Oakton VA 22124-9005 (M)
Washington Academy of Sciences
49
MORRIS, P.E., ALAN (Dr.) 4550 N. Park Ave. #104, Chevy Chase MD 20815 (EF)
MOUNTAIN, RAYMOND D. (Dr.) 701 King Farm Blvd #327, Rockville MD
20850 (F)
MUMMA, MICF1AEL J. (Dr.) 210 Glen Oban Drive, Arnold MD 21012 (F)
MURDOCH, WALLACE P. (Dr.) 65 Magaw Avenue, Carlisle PA 17015 (EF)
NORRIS, KARL H. (Mr.) 1 1204 Montgomery Road, Beltsville MD 20705 (EF)
O'HARE, JOHN J. (Dr.) 108 Rutland Blvd, West Palm Beach FL 33405-5057 (EF)
OHRINGER, LEE (Mr.) 5014 Rodman Road, Bethesda MD 20816 (EF)
ORDWAY, FRED (Dr.) 5205 Elsmere Avenue, Bethesda MD 20814-5732 (EF)
OTT, WILLIAM R (Dr.) 19125 N. Pike Creek Place, Montgomery Village MD
20886 (EF)
PAJER, BERNADETTE (Mrs.) 25116 143rd St. SE, Monroe WA 98272 (M)
PARIS, ANTONIO (Mr.) 25066 12th Ave N, St Petersburg FL 33713-5510 (M)
PARR, ALBERT C (Dr.) 2656 SW Eastwood Avenue, Gresham OR 97080-9477 (F)
PATEL, D. G. (Dr.) 1 1403 Crownwood Lane, Rockville MD 20850 (F)
PAULONIS, JOHN (Mr.) Yonkers NY 10710 (M)
PAULONIS, JOHN J (Mr.) P.O. Box 335, Yonkers NY 10710 (M)
PAZ, ELVIRA L. (Dr.) 172 Cook Hill Road, Wallingford CT 06492 (ELF)
PICKFJOLTZ, RAYMOND L. (Dr.) 3613 Glenbrook Road, Fairfax VA 22031-3210
(EF)
PLESCIA, JEFFREY (Dr.) Applied Physics Laboratory, The Johns Hopkins
University, MS 200-W230 11100 Johns Hopkins Road, Laurel MD 20723-6099
(M)
POLAVARAPU, MURTY 10416 Hunter Ridge Dr., Oakton VA 22124 (LF)
POLINSKi, ROMUALD (Mr.) 01-201 WARSZAWA, UL., Wolska 43, Poland (M)
PRZYTYCKI, JOZEF M. (Prof.) 10005 Broad St, Bethesda MD 20814 (F)
PYKE, JR, THOMAS N. (Mr.) 4887 N. 35th Road, Arlington VA 22207 (F)
RADER, CHARLES A. (Mr.) 1101 Paca Drive, Edgewater MD 21037 (EF)
RAMAKER, DAVID E. (Dr.) 6943 Essex Avenue, Springfield VA 22150 (EF)
RAVITSKY, CHARLES (Mr.) 37129 Village 37, Camarillo CA 93012 (EF)
READER, JOSEPH (Dr.) National Institute of Standards and Technology, 100
Bureau Drive, MS 8422, Gaithersburg MD 20899-8422 (F)
REAGAN, ANN M. (Dr.) PO Box 22, Lusby Maryland 20657 (M)
REDISH, EDWARD F. (Prof.) 6820 Winterberry Lane, Bethesda MD 20817 (F)
REISCHAUER, ROBERT (Dr.) 5509 Mohican Rd, Bethesda MD 20816 (EF)
RENAUD, PHILIP (Capt.) Living Oceans Foundation, 8181 Professional Place Suite
215, Landover MD 20785 (M)
RICH, PAUL (Dr.) 1527 New Hampshire Avenue, NW, Washington DC 20036 (M)
RICKER, RICHARD (Dr.) 12809 Talley Ln, Darnestown MD 20878-6108 (F)
RIDGELL, MARY P.O. Box 133, 48073 Mattapany Road, St. Mary's City MD
20686-0133 (LM)
ROBERTS, SUSAN (Dr.) Ocean Studies Board, Keck 607, National Research
Council, 500 Fifth Street, NW, Washington DC 20001 (F)
ROGERS, KENNETH (Dr.) 355 Fellowship Circle, Gaithersburg MD 20877 (LM)
Winter 2015
50
ROMAN, NANCY GRACE (Dr.) 4620 North Park Avenue Apt. 306W, Chevy
Chase MD 20815 (F)
ROOD, SALLY A (Dr.) PO Box 12093, Arlington VA 22219 (F)
ROSENBLATT, JOAN R. (Dr.) 701 King Farm Blvd, Apt 630, Rockville MD 20850
(EF)
SAUBERMAN, P.E., HARRY R (Mr.) 8810 Sandy Ridge Ct., Fairfax VA 22031
(M)
SCHINDLER, ALBERT I. (Dr.) 6615 Sulky Lane, Rockville MD 20852 (EF)
SCHMEIDLER, NEAL F. (Mr.) 7218 Hadlow Drive, Springfield VA 22152 (F)
SCHNEPFE, MARIAN M. (Dr.) Potomac Towers, Apt. 640, 2001 N. Adams Street,
Arlington VA 22201 (EF)
SEVERINSKY, ALEX J. (Dr.) 4707 Foxhall Cres NW, Washington DC 20007-1064
(EM)
SHAFRIN, ELAINE G. (Mrs.) 8168 Connecticut Ave NW Apt 2014, Washington
DC 20815-817 (EF)
SHETLER, STANWYN G. (Dr.) 142 E Meadowland Ln, Sterling VA 20164-1144
(EF)
SHIELDS, EDWARD (Dr.) PO Box 165, Grand Portage Minnesota 55605 (M)
SHROPSHIRE, JR, W. (Dr.) Apt. 426, 300 Westminster Canterbury Dr., Winchester
VA 22603 (LF)
SILBER, RONNIE (Dr.) 13710 Colgate Way #1338, Silver Spring MD 20904 (M)
SIMMS, JAMES ROBERT (Mr.) 9405 Elizabeth Ct., Fulton MD 20759 (M)
SMITH, THOMAS E. (Dr.) 3148 Gracefield Rd Apt 215, Silver Spring MD 20904-
5863 (LF)
SNIECKUS, MARY (Ms.) 1700, Dublin Dr., Silver Spring MD 20902 (M)
SODERBERG, DAVID L. (Mr.) 403 West Side Dr. Apt. 102, Gaithersburg MD
20878 (M)
SOLAND, RICHARD M. (Dr.) 2516 Arizona Av Apt 6, Santa Monica CA 90404-
1426 (LF)
SOZER, AMANDA (Dr.) 4707B Eisenhower Ave, Alexandria VA 22304 (M)
SPILHAUS, JR, A.F. (Dr.) 10900 Picasso Lane, Potomac MD 20854 (EM)
SRIRAM, RAM DUVVURU (Dr.) 1030 Castlefield Street, Ellicott City MD 21042
(LF)
STEIN, DAVID E. (Mr.) P.O. Box 571433, Las Vegas NV 89157 (M)
STERN, KURT H. (Dr.) 103 Grant Avenue, Takoma Park MD 20912-4328 (EF)
STEWART, PETER (Prof.) 417 7th Street NE, Washington DC 20002 (M)
STIEF, LOUIS J. (Dr.) 332 N St., SW., Washington DC 20024-2904 (EF)
STOMBLER, ROBIN (Ms.) Auburn Health Strategies, 3519 South Four Mile Run
Dr., Arlington VA 22206 (M)
STRAUSS, SIMON W. (Dr.) 3160 Bracefield Rd. RC 1305, Silver Spring MD
20904 (LF)
SYED, ALI (Dr) 603, H Street SW, Washington DC 20024 (M)
TABOR, HERBERT (Dr.) NIDDK, LBP, Bldg 8, Rm 223, National Institutes of
Health, Bethesda MD 20892-0830 (M)
Washington Academy of Sciences
51
TAYSING-LARA, MONICA (Ms.) 3343 Deni Place NW, Washington DC 20007
(M)
TEICH, ALBERT H. (Dr.) PO Box 309, Garrett Park MD 20896 (EF)
THEOFANOS, MARY FRANCES (Ms.) 7241 Antares Drive, Gaithersburg MD
20879 (M)
THOMPSON, DENNIS (Mr.) 325 Sandy Ridge Rd, Fredricksburg VA 22405 (M)
THOMPSON, CHRISTIAN F. (Dr.) 278 Palm Island Way, Ponte Vedra FL 32081
(LF)
TIMASHEV, SVIATOSLAV (SLAVA) A. (Dr. Prof.) 3306 Potterton Dr., Falls
Church V A 22044-1603 (F)
TOUWAIDE, ALAIN PO Box 25499, Washington DC 20027 (LF)
TOWNSEND, MARJORIE R. (Mrs.) 3529 Tilden Street, NW, Washington DC
20008-3194 (LF)
TROXLER, G.W. (Dr.) PO Box 1 144, Chincoteague VA 23336-9144 (F)
TYLER, PAUL E. (Dr.) 1023 Rocky Point Ct. N.E., Albuquerque NM 87123-1944
(EF)
UBELAKER, DOUGLAS H. (Dr.) Dept, of Anthropology, National Museum of
Natural History, Smithsonian Institution, Washington DC 20560-01 12 (F)
UMPLEBY, STUART (Professor) The George Washington University, Apt 1207
4141 N Henderson Rd, Arlington VA 22203 (F)
VAISHNAV, MARIANNE P. (Ms.) P.O. Box 2129, Gaithersburg MD 20879 (LF)
VAN TUYL, ANDREW (Dr.) 3618 Littledale Road, Apt. 203, Kensington MD
20895-3434 (EF)
VARADI, PETER F. (Dr.) Apartment 1606W, 4620 North Park Avenue, Chevy
Chase MD 20815-7507 (EF)
VAVRICK, DANIEL J. (Dr.) 10314 Kupperton Court, Fredericksburg VA 22408
(F)
VIZAS, CHRISTOPHER (Dr.) 504 East Capitol Street, NE, Washington DC 20003
(M)
WALDMANN, THOMAS A. (Dr.) 3910 Rickover Road, Silver Spring MD 20902
(F)
WALLER, JOHN D. (Dr.) 5943 Kelley Court, Alexandria VA 22312-3032 (M)
WEAR, DOUGLAS (Dr.) 8014 Barron Street, Takoma Park MD 20912 (M)
WEBB, RALPH E. (Dr.) 21-P Ridge Road, Greenbelt MD 20770 (EF)
WEIL, TIMOTHY (Mr.) SECURITYFEEDS, PO Box 18385, Denver CO 80218
(M)
WEISS, ARM AND B. (Dr.) 6516 Truman Lane, Falls Church VA 22043 (LF)
WERGIN, WILLIAM P. (Dr.) 1 Arch Place #322, Gaithersburg MD 20878 (EF)
WHITE, CARTER (Dr.) 12160 forest Hill Rd, Waynesboro PA 17268 (F)
WIESE, WOLFGANG L. (Dr.) 8229 Stone Trail Drive, Bethesda MD 20817 (EF)
WILCOX, ROBIN J. (Ms.) 8601 Park Avenue, Bowie MD 20720 (M)
WILLIAMS, CARL (Dr.) 2272 Dunster Lane, Potomac MD 29854 (F)
WILLIAMS, E. EUGENE (Dr.) Dept, of Biological Sciences, Salisbury University,
1101 Camden Ave, Salisbury MD 21801 (M)
Winter 2015
52
WITHERSPOON, F. DOUGLAS ASTI, 11316 Smoke Rise Ct., Fairfax Station VA
22039 (M)
WU, KLELI (Mr.) 360, Suite 48, Swift Ave, South San Francisco CA 94080 (M)
Washington Academy of Sciences
53
Washington Academy of Sciences
1200 New York Avenue
Rm 113
Washington, DC 20005
Please fill in the blanks and send your application to the address above. We will
contact you as soon as your application has been reviewed by the Membership
Committee. Thank you for your interest in the Washington Academy of Sciences.
(Dr. Mrs. Mr. Ms)
Business Address
Home Address
Email
Phone ===__
Cell Phone _^__________=======_===__^=__=
preferred mailing address Type of membership
Business Home Regular Student
Present Occupation or Professional Position
Please list memberships in scientific societies - include office held:
Winter 2015
54
Instructions to Authors
1. Deadlines for quarterly submissions are:
Spring - February 1 Fall - August 1
Summer - May 1 Winter - November 1
2. Draft Manuscripts using a word processing program (such as
MSWord), not PDF. We do not accept PDF manuscripts.
3. Papers should be 6,000 words or fewer. If there are 7 or more graphics,
reduce the number of words by 500 for each graphic.
4. Include an abstract of 150-200 words.
5. Include a two to three sentence bio of the authors.
6. Graphics must be in graytone, and be easily resizable by the editors to
fit the Journal’s page size. Reference the graphic in the text.
7. Use endnotes or footnotes. The bibliography may be in a style
considered standard for the discipline or professional field represented
by the paper.
8. Submit papers as email attachments to the editor or to
wasioumal@washacadsci.org .
9. Include the author’s name, affiliation, and contact information -
including postal address. Membership in an Academy-affiliated society
may also be noted. It is not required.
10. Manuscripts are peer reviewed and become the property of the
Washington Academy of Sciences.
1 1 . There are no page charges.
12. Manuscripts can be accepted by any of the Board of Discipline Editors.
Washington Academy of Sciences
Washington Academy of Sciences
Affiliated Institutions
National Institute for Standards & Technology (NIST)
Meadowlark Botanical Gardens
The John W. Kluge Center of the Library of Congress
Potomac Overlook Regional Park
Koshland Science Museum
American Registry of Pathology
Living Oceans Foundation
National Rural Electric Cooperative Association (NRECA)
Winter 2015
56
Delegates to the Washington Academy of Sciences
Representing Affiliated Scientific Societies
Acoustical Society of America, Washington Chapter
American Assoc, of Physics Teachers, Chesapeake Section
American Astronomical Society
American Ceramics Society
American Fisheries Society
American Institute of Aeronautics and Astronautics
American Meteorological Society, Washington, DC Chapter
American Nuclear Society, Washington DC Section
American Phytopathological Society, Potomac Division
American Society for Cybernetics
American Society for Metals, Washington Chapter
American Society of Civil Engineers, National Capital
Section
American Society of Mechanical Engineers, Washington
Section
American Society of Microbiology, Washington Branch
American Society of Plant Biologists, Mid-Atlantic
Anthropological Society of Washington
ASM International
Association for Women in Science, DC Metropolitan Chapter
Association for Computing Machinery, DC Area Chapter
Association for Science, Technology, and Innovation
Association of Information Technology Professionals
Biological Society of Washington
Botanical Society of Washington
Capital Area Food Protection Association
Chemical Society of Washington
District of Columbia Institute of Chemists
District of Columbia Psychological Association
Eastern Sociological Society
Electrochemical Society, National Capital Section
Entomological Society of Washington
Geological Society of Washington
Flistorical Society of Washington DC
Human Factors and Ergonomics Society, Potomac Chapter
Institute of Electrical and Electronics Engineers, Northern
Virginia Section
(continued on next page)
Paul Arveson
Frank R. Haig, S. J.
Sethanne Howard
Vacant
Lee Benaka
David W. Brandt
Vacant
Charles Martin
Vacant
Stuart Umpleby
Vacant
Vacant
Daniel J. Vavrick
Vacant
Mark Holland
Vacant
Toni Marechaux
Jodi Wesemann
Alan Ford
F. Douglas
Witherspoon
Chuck Lowe
Stephen Gardiner
Chris Puttock
Keith Lempel
Elise Ann Brown
Vacant
Tony Jimenez
Ronald W.
Mandersheid
Vacant
Vacant
Jeffrey B. Plescia
Jurate Landwehr
Vacant
Gerald P. Krueger
Murty Polavarapu
Washington Academy of Sciences
Delegates to the Washington Academy of Sciences
Representing Affiliated Scientific Societies
(continued from previous page)
Institute of Electrical and Electronics Engineers, Washington
Section
Institute of Food Technologies, Washington DC Section
Institute of Industrial Engineers, National Capital Chapter
International Association for Dental Research, American
Section
International Society for the Systems Sciences
International Society of Automation, Baltimore Washington
Section
Instrument Society of America
Marine Technology Society
Maryland Native Plant Society
Mathematical Association of America, Maryland-District of
Columbia- Virginia Section
Medical Society of the District of Columbia
National Capital Area Skeptics
National Capital Astronomers
National Geographic Society
Optical Society of America, National Capital Section
Pest Science Society of America
Philosophical Society of Washington
Society for Experimental Biology and Medicine
Society of American Foresters, National Capital Society
Society of American Military Engineers, Washington DC
Post
Society of Manufacturing Engineers, Washington DC
Chapter
Society of Mining, Metallurgy, and Exploration, Inc.,
Washington DC Section
Soil and Water Conservation Society, National Capital
Chapter
Technology Transfer Society, Washington Area Chapter
Virginia Native Plant Society, Potowmack Chapter
Washington DC Chapter of the Institute for Operations
Research and the Management Sciences (WINFORMS)
Washington Evolutionary Systems Society
Washington Elistory of Science Club
Washington Paint Technology Group
Washington Society of Engineers
Washington Society for the History of Medicine
Washington Statistical Society
World Future Society, National Capital Region Chapter
Richard Hill
Vacant
Neal F. Schmeidler
J. Terrell Hoffeld
Vacant
Vacant
Hank Hegner
Jake Sobin
Vacant
D. S. Joseph
Vacant
Vacant
Jay H. Miller
Vacant
James Cole
Vacant
Eugenie Mielczarek
Vacant
Daina Apple
Vacant
Vacant
E. Lee Bray
Terrell Erickson
Richard Leshuk
Vacant
Russell Wooten
Vacant
Albert G. Gluckman
Vacant
Alvin Reiner
Alain Touwaide
Michael P. Cohen
Jim Honig
Washington Academy of Sciences
Room 113
1200 New York Ave. NW
Washington, DC 20005
Return Postage Guaranteed
NONPROFIT ORG
US POSTAGE PAID
MERRIFIELD VA 22081
PERMIT# 888
*
CMA'/ft/- :
•' ! C? O 1 \ / J