Previous Final Projects
- Battling the Binge: A survey of eating disorders at
Colgate, with results compared to national trends.
- Breaking the Bubble: A survey of Colgate students
to learn what sources they use to keep up with national and
international news, and how informed they feel about issues
beyond the campus. The survey had to be designed carefully
in order that the results were numeric and hence could be
- Stereotypes at Colgate -- a Collective Statistical Study:
Colgate students were presented with yearbook descriptions of
high school students and asked to guess which were members of
different minority groups. But the methods used could not have
produced much useful information.
- National Hockey League Money and Performance: Relating
several ways of measuring quality of players and teams (on the
basis of game statistics) and their associations with pay levels.
Sports data is relatively easy (though not trivial) to obtain,
but since "quality" is subjective in this context, a project like
this must include clear arguments to establish why the
particular statistics used here were particularly good measurements
- Correlations between Achievement Orientation and Starter
vs. Non-Starter Athletes: Analysis of a psychological survey
to find differences between the motivation and mind-set of
women soccer and field hockey players who start a game and those
who do not. There is plenty of opportunity in this project to
comment on the difference between correlation and causation and
how things might have been done differently -- but that should
be a big part of every project, and very few
do it well.
- A Statistical Look at Colgate's E-mail System:
The computer center kindly provided numbers of e-mail messages
sent during various time periods, and these were presented in
the form of bar charts. Limited time for data gathering made
- The Chance of Getting Hurt Tomorrow: Analysis of data
on football injuries sustained while playing on grass vs.
- The Effects on the Three-Point Rule Changes
(1994-95 season) on the Top Players in the NBA: Deciding
whether rule changes intended to make more 3-point goals
in basketball easier were effective in allowing more scoring.
Though the project was done in '97, data was available only
for the top scorers (and was missing for some seasons even
in these cases), so comparisons were difficult.
- AIDS and Zidovudine: Does Zidovudine Slow the Progression
of AIDS? The source of the data was a table in an article
in The New England Journal of Medicine. The authors of
that article did their own statistical study, using methods
beyond the scope of our course, but the author of this project
used methods we had. She was limited by the fact that
the table contained only a summary of the experimental data.
- The Role of Anxiety, Depression, Hostility and Stress
in Type A Versus Type B Personality and Vulnerablility to
Coronary Heart Disease: Reporting on a psychology project
which used methods not thoroughly covered in our course
(ANOVA). It is tempting, in reporting on projects done in
other departments, to focus on the subject matter of that
department. But the final project grade is based on the
presentation of the statistics: After making sure that the
experiment is clearly explained to someone not specializing
in that area, the project should give a thorough explanation
of (1) the choice of which statistics to compute, (2) the
method used to compute them, if it is not covered in our class,
and (3) the conclusions drawn from them. This should not be
a journal article in the other field, but a project on
- Is the National Crime Rate Decreasing?: Department
of Justice figures, obtained from the Internet, on violent
crime suggest that fewer crimes per person are occurring.
The Internet data were only totals, however, so many questions
(e.g., are burglaries decreasing in affluent neighborhoods?)
had to be left unstudied.
- Media, China and Public Opinion: A survey of students
to tell whether their responses to various statements about
China were influenced (positively or negatively) by having
taken courses on China or by getting all their information
on China from the mass media. Limited sample size, especially
of students who had taken courses on China, proved to be a
- Relationship of Water Nutrients and Zooplankton to
Pond Phytoplankton in Central New York: Reporting results
on an ecology project. See "The Role of Anxiety, ..." above.
- Stroop Interference While Evaluating Same or Different
Ink Colors of Words and Shapes: Students were shown flash
cards with the same word or different words written in the same
color or different colors and asked if they were the same or
different word; or the word RED written in green, etc., and
asked to read the word. Differences in comprehension/reaction
times were analyzed. Some of the analysis of the data was done
with methods within the scope of our course, while some
required ANOVA. See "The Role of Anxiety, ..." above.
- 1997 Baseball Statistics versus 1997 Baseball
Salaries: Are the top players overpaid? Are they paid in
proportion to their production? Data was available on the
Internet only for top players, so the data set was small and
comparisons to "journeymen" could not made.
- A Statistical Glance at U.S. News and World Report's
1998 College Rankings: What is the Difference Between a National
University and a Liberal Arts College Education?: Study of
the data that goes into the U.S. News rankings (some of
which is available only in Peterson's Guide). As the
authors pointed out, many factors that go into choice of
college or university could not be quantified and hence could
not be analyzed statistically.
- Stats Regression Project: Determining What Makes an NBA
Team Win: Stats from a pro basketball handbook.
- A Study of Marijuana and Cigarette Use on "The Hill"
at Colgate: Presentation in several forms of data from a
survey, to determine levels of use in the residence
halls, addictiveness, and whether nicotine is a "gateway drug"
for marijuana. Very deep matters for a 5-question survey on
confidential matters, so the results were very questionable;
but since they were numeric, they were easily accessible to
statistical analysis. Here is a great opportunity for comment
on "what I would have done differently if ..."
- A Test for Correlation Between Country Position and Per
Capita Income in the Developing Nations of the World:
Does distance from the Equator (or the resulting climate)
affect per capita income? Data from an atlas. But how is
distance from the equator measured for a large country?
- Is There a Home Field Advantage in the World Series?:
Comparing wins versus losses at home and on the road in World
Series games. Should pure numbers have been used? Might it be
possible to include data on the eventual winner, or the teams'
records before the series? If so, how?
- Does a Background or Interest in Art and Art History
Allow for a More Accurate Assessment of the Value of a
Painting?: Surveying students: "How much would you pay for
this painting [done by a master], or for this painting [in a
similar style, done by the author of the project]?"
Revised January 5, 2000. Questions to
Copyright 2000 © Colgate University. All rights reserved.