Robert Fowler
13 May 2003
Overview
The Statistical Consulting Center at Florida State University is a
research assistance facility for the students, faculty, and staff at FSU.
Occasionally, clients from outside of FSU call and are given a one-time,
one-hour appointment. The Statistical Consulting Center is a free
service for members of the FSU community. Services included but are
not limited to:
Translating hypotheses into statistical terms
Designing sampling procedures
Choosing appropriate statistical methods
Interpreting computer output
Phrasing statistical results
Referrals to other statistical help.
The Statistical Consulting Center generally does not perform actual
analyses.
Appointments for the Spring Semester of 2003 were scheduled from 11:00
a.m. 1:00 p.m. on Monday, Wednesday and Friday. Alternative arrangements
are made when clients are unavailable during the scheduled hours.
During the Spring semester of 2003 everyone who requested assistance was
seen by the Statistical Consulting Center within 9 business days of his
or her initial call for assistance.
Summary of Business Activities
The Statistical Consulting Center changed management on 13 January 2003.
Feiming Chen managed the Statistical Consulting Center prior to January
2003. He saw 56 clients from May 2002 to December 2002. Historically
Consulting Center clients have been approximately 85% graduate students,
10% faculty, and 5% undergraduates and clients from outside FSU.
The Statistical Consulting Center saw 26 individuals from 13 January
2003 to 1 May 2003 and handled a half-dozen brief questions by phone or
email. The 26 clients were comprised of 4 faculty members, 2 clients
from outside FSU, 17 graduates students working on their theses or dissertations,
and 3 graduate students working on publishable research related to their
dissertation topic. Additionally, two undergraduates accompanied
one of the faculty members had questions related to the faculty member’s
research.
The clients from the Spring of 2003 came from across the University
from a variety of departments including:
Anthropology
Biology
Chemistry
Educational Psychology and Learning Systems Department
Family and Children Sciences
Family and Consumer Sciences
Mathematics Education
Oceanography
Physical Education
Physics
Sports Management and Recreation Management
Textiles and Consumer Sciences
Clients also included a doctor from the Tallahassee area and a recent
Ph.D. graduate now working in Miami.
Generally each client was seen twice. Many of the questions involved
designing and understanding appropriate sampling designs and analysis methods.
These issues included propagation of errors, reliability, validity, sample
size, and power. More than half the clients had data and need assistance
in either choosing the appropriate methods or interpreting results.
There were many questions about the operation of SPSS, and some other software
packages. Many of these questions I was able to answer. For
questions regarding packages I was unfamiliar with or for highly specific
questions with software that I am familiar with several clients were referred
to Betty Brown, the computational applications specialist at ACNS.
Analyses were performed for two professors to better help them understand
the methods being suggested. In one of these cases the analysis was
performed for Dr. James in Biology, a long-term collaborator of Dr. Meeter.
The analysis for Dr. James confirmed the analysis done by Dr. Chuck McCulloch,
from the University of California at San Francisco. The second analysis
was done for Dr. Behzad Mortazavi in Oceanography. Dr. Mortazavi
will be acknowledging the Statistical Consulting Center in an article that
has now been approved for publication.
Clients of the Statistical Consulting Center have a large variety of
statistical backgrounds. Some clients were unfamiliar with the basic
terminology and concepts of statistics. Other clients had a substantial
understanding of statistical concepts and techniques and required assistance
on topics that required me to learn additional background information in
order to assist them.
Typical Cases
The cases during the Spring of 2003 varied from highly technical to
clients satisfied after a three-minute phone conversation and a one-sentence
answer. The typical case involved two appointments for a combined
meeting time of between three to four hours. Most clients had either
already designed a sampling device and they were in the process of collecting
data or had already collected data. Most want to insure they had
collected appropriate data and obtain guidance on what functions to use
in SPSS to do an appropriate analysis. Generally after the first
meeting there were several issues for the client to work out before running
the analysis. After running the analysis clients return to help understand
the computer output and translate the statistical results into language
that worked for their thesis or dissertation.
A case in point is that of a recent Ph.D. graduate. She is working
with doctors in Miami to determine if there is a difference in the pain
thresholds between various types of migraine headache patients. In
our first meeting we discussed the possible problems of using 16 factors
in a regression analysis type of problem. Additionally, she had data
from both sides of each patient (left and right) and wanted to compare
the left and right sides when the hypotheses the doctors were interested
in involve comparing the pain on the same side of the body as the headache
and the opposite side. We discussed ways to bring her analysis into
agreement with the ideas the doctors were interested in. Two weeks
later she returned with her new analysis and we walked through the consequences
of the output. When we concluded our meeting she was comfortable
that she could correctly explain the results of her analysis to the doctors.
Reflections
Working in the Statistical Consulting Center has been an exciting experience.
The clients bring real problems without the niceties of textbook problems.
Every client has taught me something new, usually about a field of study
I barely new knew existed before the meeting. I look forward to continuing
the service that the Statistical Consulting Center provides for the FSU
community.
I would like to acknowledge Dr. Lin, Dr. Song, and Dr. Zahn for their
help with several of the cases. I would also like to thank Virginia
Hellman, Pam McGhee, Evangelous Robinson,
Delenie Garrido, James Stricherz, and Catherine Lingar. They
have kept the operations of the Statistical Consulting Center running smoothly.
|
