UW Undergraduate Research Program

<<< URP Home

Students in Research
See more URP photos!

URP Advising

Creative Commons License
The Undergraduate Research Program website, created by the Undergraduate Research Program at the University of Washington, is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.

Permissions beyond the scope of this license are available at exp.washington.edu/urp/about/rights.html

The Washington Research Foundation Fellowship

Student Experiences

Jeff Eaton - Sociology, Statistics, Mathematics

Jeff's Project - Bayesian Melding to Improve Parameter Estimation in Stochastic HIV Epidemic Modeling

Mathematical models of epidemics are used my epidemiologists, doctors, and policy makers to distill complex disease phenomena into component parts in order to understand and predict what may happen in a variety of scenarios. This information is useful for making informed decisions about controlling and intervening in epidemics, such as the global HIV epidemic. For example, a government may use a model to predict the amount anti-retroviral drugs that would be needed to treat all those living with AIDS in a country, or an epidemiologist may use a model to understand why or why not a particular intervention strategy is effective. Microsimulation models are models that simulate specific individuals in a population by exposing them to random hazard of important events, such as infection, marriage, or death, in each time step. Because of the ability to specifically model complex individual behavior, microsimulation models are very useful for understanding intricate social dynamics. However, it is very difficult to accurately quantize individuals' behavior and there is a severe lack of high quality population data in the parts of the world worst afflicted by infectious diseases such as HIV, so microsimulation models often can only be used to obtain theoretical results about hypothetical populations. Recently concerted efforts have been made to improve data collection in Africa, and statistical techniques have been developed for estimating parameters even in the case of imperfect data. Combining these it may be possible to develop microsimulation models that recreate actual populations.

When, how, & why did you get involved in research?

For me, undergraduate research was an opportinity to be much more active in my education. My first experience with undergraduate research were during my sophomore year in which I participated in I participated in mathematics research on discrete electrical networks with Professor Jim Morrow in an NSF funded REU program. I also worked on a large educatoinal survey project run by Sociology Professor Charles Hirschman called the 'Beyond Highschool Project'. These short research showed me both the excitement and frustration of trying to solve problems for which the answer is not already known. My introduction to epidemiological modeling in a course on Demographic Methods taught by Sociology Professor Dr. Samuel Clark. I find epidemiolical modeling particularily stimulating because of the interdisciplinary nature of the work. Deep knowledge of biology, sociology, mathematics, and statistics must all be applied in order to develop an insightful model. After my enthusiasm in the course, Professor Clark offered to mentor me as an undergraduate research assistant. Working together we have developed a model to analyze the possible impact of male circumcision as an intervention strategy in the HIV epidemic.

For me the most exciting aspect of my research is the opportunity to share it with other people. Because of the enormous policy implications of epidemiological research and the sometimes tenuous assumptions involved in a model, discussing modeling results can be intense. Often a breif conversation, formal or informal, can point out an idea that may never have occurred and stimulate months of further work.

The most frustrating aspect of my undergraduate research experience is simply when things don't work. The key difference between research and exercises for courses is that for research questions the solution is not known, and usually there is not an obvious path toward the solution. I've realized that I have a tendency to severely underestimate the amount of time and energy that it will take to accomplish tasks. While this proves to be frustrating when things aren't going well, I believe that it is also what provides satisfaction when a solution is finally eminent.

What advice would you give to other student?

In general, my advice about getting involved in undergraduate research is to ask a lot of questions. Asking questions will guide you towards your interests. Find a professor in whom you are interested and go to his or her office hours. This is the place where you can find out in what he or she is really interested. In my "field" (in quotations because I have a hard time labling exactly what my field is yet), I think that it is important to develop a broad knowledge of scientific disciplines, and the ability search for and learn when you need a specific peice of knowledge. Communication skills between individuals of differing expertise is essential.

Future Plans

After my undergraduate career I would like to pursue graduate studies in a field related to epidemiological modeling and work as a consultant for international policy making institutions.


I'd like to sincerely thank Sociology Professor Samuel Clark. Professor Clark has dedicated an enormous amount of time developing my research skills, introducing me to the field of demographic and epidemiological modeling, and advising my research project, as well as generally mentoring me as I make decisions that will affect my career.

Awards and Honors

  • Mary Gates Undergraduate Research Scholarship
  • Barry M. Goldwater scholarship
  • Mary Gates Endowmnent Venture Fellowship
  • College of Arts & Sciences Research Award
  • Research Fellowship for Advanced Undergraduates


  • Center for Statistics & Social Sciences Student Seminar
  • UW Mathematical Modeling Group
  • UW Undergraduate Research Symposium
  • TB & HIV Modeling Conference