In my last two posts (here and here), I’ve talked through the rationale for undergraduates doing replication and shared a replication assignment of my own. In the final segment of this series, I want to talk a little bit about strategies for developing your own successful replication assignment.
First, start with a highly readable article. Alas, this means that most of what is in the APSR is out. Both Perspectives on Politics and PS: Political Science and Politics have appropriate empirical articles, though, that are typically shorter, more accessible, and less technically sophisticated. The better students can understand the article, the better their chances of success on the assignment. Even if you don’t usually give reading guides, I’d consider doing one (just reading comprehension questions for students to review while reading) for this assignment to help ensure everyone starts off on the right foot with a good understanding of the reading.
Second, talk it up in class. Emphasize that research gets published using only the skills the students currently have, and we’re going to show that to ourselves by replicating published research. This is something they should be proud to be able to do; it’s an achievement.
Third, consider allowing students to work with partners or in trios, even if they turn in separate written work. Working together will give them – especially the women – more confidence about their ability to do the tasks, and that will reduce both the stress on them and the number of anxious questions you’ll get.
Finally, give yourself plenty of time to write the assignment. Be detailed and specific. It will also take you some time to get and clean the data, and possibly write a sketchy codebook in the instructions, so that the assignment is plug-and-play ready when it goes out to students. You’ll want to drop most of the unused variables (especially if there are fixed effects dummies you aren’t using) and similar clutter like that. There’s a reason I’m posting this blog entry now: writing one of these would be an excellent summer project, a good activity for when you’re stalled on your research or just need to change mental gears for a bit. This will take a few hours, but like all good problem sets, once you’ve written it you can reuse it repeatedly.