A Summer Course in Statistics

        By Eric L. "Skip" Krueger, Ph.D. student, Department of Political Science (regress@gte.net)

        Everyone in the Political Science Department was very excited that I was going to be attending the summer course in statistics at the University of Michigan last summer, a trip sponsored in part by the UNT Computing Center. But I must admit to a little doubt - a lot was going on, and I was already beat by July. My wife summed it up the chaos that was my summer: "We just got married and you<@146>re leaving me for a month?"

        "Well, yeah. So, off I trudged to Ann Arbor, tired, grumpy and completely unprepared for six hours a day of statistics classes. Man, was I wrong. While it's true that I missed my wife, the time passed quickly. The readings were (usually) interesting enough to keep me awake even after six hours of classes. And the faculty - selected from political science, sociology and economics departments across the country for their expertise - were extremely generous with their time. I left perhaps more energized than educated, but it was just what I needed as I started last semester preparing for comps and my dissertation.

        A number of professors in Political Science had been to the program, to which any UNT graduate student can apply, or they knew the people who were going to be teaching there, and they were very free with their advice. Despite all the useful advice - advice anyone contemplating attending the program should seek out (not all the teachers were awesome, for example) - I was completely unprepared in some areas, so let me drop just a few tips for those interested.
         

        • My first and best advice is to con a friend - especially one studying in your substantive field - into going with you. During the first week, students introduced themselves to each other - maybe - but I had a hard time finding other political science students. I quickly gave up on finding people interested in international political economy. In 1996, I hear, it was all political scientists - no sociologists or economists. So, the composition changes year to year.
         
          Even after you start meeting people, however, you don't really hang out with them a whole lot. Unlike the European counterpart at the University of Essex, the Michigan program does not house students together in dorms or other campus housing. Not only do students have to find their own housing (no small pain), they are spread all around campus. There is not central bar, restaurant, or lounge where students tend to congregate to relax and socialize. This is particularly strange since all the UNT faculty felt that the social aspects of attending the program were significantly important (after all, I will be working along side some of these people in a couple of years).
           
        • You would also like to be able to share expenses. To attend as a non-credit student (the cheaper way to go) will cost at least $1,500 plus travel (say, another $300). You could significantly cut that expense by sharing an apartment with somebody from UNT. (Forget about finding a cheap sublet in Ann Arbor - it's impossible. In fact, finding anything cheap in Ann Arbor is impossible.)
         
        • You should know exactly what you want to learn before you go. In other words, do your homework. Classes there are not like classes at UNT. As a non-credit student (or so-called "visiting scholar") you needn't worry about tests. This lifts a burden off most students (some are taking the classes for credit - poor sods), and it allows them to focus on those aspects of the course that interest them most. But it also means that there are way more things to chose from than you have time for. Also remember, don't waste precious time and money taking things there that you can get at UNT, whether in your home department or another. I couldn't believe, for example, how many people from big-name schools were sitting in the introductory regression class.
         
        • Finally, bring your own data sets. The program's brochure says no facilities are available for analysis of one's own data. Bull. In fact, most of the teachers encourage you to try out the fancy new techniques you're learning on your own data. Oh, and having a problem getting your data to run? No prob. - just stop by the professor's office any time and he or she will help you fix it. Take some time series data, take some panel data (pooled cross-sectional time series, as we political scientists say), take some count (or duration) data. So the program's computer labs are crowded in the afternoons - so are the ones at UNT. The labs are open late, and there's not much else to do.

          • For those interested in attending, you can get information from Karl Ho in the UNT Computing Center (kho@unt.edu). Feel free to E-mail any questions to me at regress@gte.net also.

          Previous Article <== ==> Next Article

          If you have any problems or questions about this server, contact us as soon us as soon as possible. You can send mail to the following address: www@unt.edu