# Thomas Parker

Q: Where did you go to
university?
A: “I got my undergraduate degree at the University of Illinois at Urbana-Champaign, and then I went to UBC to do a master’s degree in economics.  Then I ended up going back to Illinois for my PhD.”

Q: What area of study in economics do you specialize in?
A: “I do my research in econometric theory, but I’m open to suggestions…”

Q: What research are you currently working on?
A: “Currently I’ve been working on a few projects that all deal with roughly the same thing: how to find out more about a collection of data than what an average tells you. By average, I mean usually what a standard regression is constructed to tell you about, which is how the conditional average of a response variable depends on some explanatory variables.  My favourite way to get a more complete picture of the distribution of data is by using quantiles, which is simply the generalized term for features like quartiles, percentiles, etc.”

“Currently I’ve been working on ways to learn from many quantiles at once, for example (in a project with Pierre Chaussé), to use them to make accurate hypothesis tests.  An example of the sort of hypothesis we’re concerned with might be whether all the quantiles of one distribution are higher than the corresponding quantiles of another distribution.  If, for instance, your two distributions represented individuals who had or had not been treated in some way (like receiving a drug treatment or unemployment benefit), this hypothesis would mean that the treatment had affected all of the members of the treated group in the same way.  Making sure you can believe that the two distributions are unambiguously different is important, and we’ve been working on ways to make drawing such inferences as accurate as possible.” (like discrete choice or panel data).  Along the way, it involves inspecting the data and diagnosing potential problems, and a lot of computational tricks to get useful answers out of your silicon friend.”
I suppose you could say that about any quantitative social science work; I just happened to take economics classes at the right time of my life, and I never stopped because it all seemed so interesting.”

Q: What classes do you teach here at the university?
A: “I’ve taught ECON 221 (an introduction to statistics) before, but this term I’m doing something new: ECON 422, which is called, “Topics in Econometrics.”  This term I decided to make the course about advanced data analysis, and we’re looking at extensions of the typical regression model that you need when your data isn’t as nice as the standard regression model assumes it is (like discrete choice or panel data).  Along the way, it involves inspecting the data and diagnosing potential problems, and a lot of computational tricks to get useful answers out of your silicon friend.”

Q: And how long have you been teaching at uWaterloo?
A: “This is my third year at the University of Waterloo.”