A Q&A with graduate student Liz Pyburn


SUMMARY: Graduate School Thesis Award winner Liz Pyburn discusses why she chose JMU and mentors who shaped her experience.

Why JMU?

I received my bachelors in psychology at Liberty University in Lynchburg, VA. One of my faculty mentors in the psych program did a lot of work in psychological statistics, and after I graduated he helped me get a job in Liberty’s Institutional Effectiveness office doing data analysis work related to academic program assessment. I enjoyed the work, and after two years there decided to apply to psychology masters programs that had a quantitative and/or statistical focus. A friend of mine from Liberty had completed JMU’s MA in Psychological Sciences: Quantitative Concentration program, which is how I heard about JMU and the quant program specifically. I decided to apply, was accepted, and here I am! I completed the masters program in May of 2015, and now I’m in the Assessment and Measurement Ph.D. program at JMU.

Mentors that shaped your experience towards completing your master's thesis?

Definitely my advisor, Dr. Jeanne Horst. I was one of her very first advisees, and she is a natural advisor. She was so supportive and was open to whatever I wanted to do for research from the very beginning of the program. The other two members of my thesis committee, Dr. Dena Pastor and Dr. Monica Erbacher, were also super helpful. Dena is our resident expert on mixture modeling (one of the analyses I used for my thesis) and Monica was familiar with cluster analysis (the other analysis I used). They both helped Jeanne and troubleshoot and figure out the nuances of each technique.

What made your JMU experience special?

Definitely the people in my program. Although the quant concentration is technically part of the larger Psychological Sciences masters program, we’re fairly distinct from the other concentrations and as a result are very close-knit. The students are all very supportive of each other, both personally and professionally. We all collaborate on research projects together and help each other in classes. The faculty are also extremely supportive and approachable. I never hesitate to approach any faculty member with a question I might have, or to ask them if I can help with their research. I know students in programs at other universities can be competitive, and faculty can be standoffish if they’re busy with their own research; the fact that the program at JMU was not like that is one of the things that attracted me to it.

Details about your project? For example, what inspired your research? Favorite part about your research? Description of the process itself?

All first-year students in the Psych Sciences program have to complete a first-year research project. Jeanne, my advisor, presented me a list of research project she was involved with when I first started the masters program and asked if I would be interested in any of them as a first-year project. One of them involved applying a technique called cluster analysis to some data from the Office of International Programs. I thought cluster analysis sounded interesting, so I picked that project and ended up liking it so much that I wanted to pursue it as part of my thesis.

Some background: cluster analysis involves grouping people based on similarity of responding to a series of variables. For example, suppose you completed a personality scale to assess your levels of openness, conscientiousness, and extraversion. Also suppose you were high on openness and conscientiousness, but low extraversion. In cluster analysis, you would be grouped with respondents who displayed a similar pattern. Then there might be another group of people who were low on openness and conscientiousness, but high extraversion; and other groups with different patterns. Cluster analysis is a person-centered method, because the focus of the analysis is on the pattern of responding across variables, within people; whereas analyses we usually use in psychological research are focused on the pattern of responding across people, within variables (e.g., how do people differ on extraversion).

There’s another analysis that is similar to cluster analysis called mixture modeling. It does basically the same thing, but in a much different way. For my thesis, I compared the performance of cluster analysis to the performance of mixture modeling. To do that, I applied both techniques to a sample of undergraduate data and compared the results.

My favorite part of my research was gaining a deeper understanding of both techniques. We have computer software that can do all these analyses, but it’s easy to just run analyses without really understanding what’s going on behind the scenes. By writing about, and conducting, these analyses, I learned so much more about the analyses. I also learned a lot about the research and writing process. I was able to produce a written product that I’m really proud of, which is a nice feeling.

Back to Top

Published: Wednesday, June 15, 2016

Last Updated: Wednesday, October 25, 2017

Related Articles