Since 1963, the ICPSR summer program in Ann Arbor has trained some 21,000 participants from more than 40 countries in quantitative research methods. One of last year’s participants was Simon Davidsson, doctoral student in political science at Lund University. We called Simon to talk about the summer school, research and open data.
What is the topic of your dissertation?
- I’m researching the development of parliamentarism in Western Europe. I work with historic literature to create a dataset in order to see which actors can influence the resignation of a government, and how this changes over time. The starting point varies for different countries. For Belgium, for instance, I start in 1830 when the country was born, whereas for Great Britain I go as far back as the 1720s.
You received one of SND’s travel scholarships in 2018 and spent eight weeks in Ann Arbor. Which courses did you take there?
- During the first period I took “Measurement, Scaling and dimensional analysis”, “Regression III” and “Mathematics for Social Scientists II” during the day. I also took an evening course in order to learn the R software application. During the second period I took “Maximum likelihood estimation II” and “Empirical modeling of social science theory”. I also audited “Maximum likelihood estimation I” during the first period and “Causal Inference” during the second.
What should you bear in mind when you choose courses?
- Even though you can pick courses when you arrive in Ann Arbor, it’s good to have given some thought to what to choose between before you go, and to make sure that you don’t choose courses that are too easy. I was a bit out of practice in regression analysis, as I hadn’t worked with it in three years. But I still managed Regression III, as you get a lot of repetition along the way.
No matter which courses you pick, you should be aware that there will be a lot of work involved. I studied the R software application at night during the first two weeks, and that was hard. If I had known how many 12-hour days I would have, I would have made sure to brush up on my programming skills, so I wouldn’t have had to learn that in Ann Arbor.
Do you have any other useful information about what might be worth thinking about before you attend the summer school?
- I recommend arranging lodging as close to campus as possible. I also think that you should attend the social activities that ICPSR arrange, as it can be difficult to get to know people in the classroom. It might be a good idea to wait until your second year as a doctoral student before you apply for Ann Arbor. I had only been a doctoral student for ten months when I was there, and didn’t yet know what method to use. So for better or worse, the courses gave me more breadth than depth.
As for what to focus on, I would make these priorities: attend the lectures, do the exercises, gather code, talk to the teachers and other students, read the literature. And make sure to get some time off!
One more thing: it might be good to bring an adapter. I didn't…
What was the best thing about your time in Ann Arbor?
- First of all, you come home with lots of articles, code, software knowledge, and self-confidence. Second, and just as important, you’ve got connections to all the other doctoral students and have had discussions with and talked to the teachers in the program.
ICPSR does more than arrange the summer school; it's also an important infrastructure for disseminating research data. How much did they discuss open data in the courses?
- I didn’t really come across anything about open data, data dissemination, or data management. Perhaps they discussed it in courses that I didn’t take. We had a lecture on ICPSR and got information about their research data and how to use them. But there were no fundamental discussions on open access to data.
What is your view on open access? In your research, you have used other researchers’ data as well as collected data of your own.
- I think it’s incredibly important with open data; it is fundamental that we share. It’s a matter of transparency as well as cost. As long as the legal requirements are met, there are no reasons not to share data, at least not after some time has passed. It’s a matter of give and take, and I don’t think that sharing more data will give people fewer incentives to collect new data.
