Aims

This course assumes students have taken a year of stats/methods training and seeks to pragmatically bridge the gap between standard textbook coverage of methodological and statistical issues and the complexities of everyday behavioral science research. It focuses on issues that (i) behavioral researchers are likely to encounter as they conduct research, but (ii) may struggle to figure out independently by consulting a textbook or published article. Topics meet this second criterion for one of four reasons:

1. They are technically challenging (e.g., When, and what does it mean, to cluster standard errors? What if data severely violate a test’s assumption? How to bootstrap.)
2. There isn’t yet consensus among methodologists about them and hence a behavioral researcher will encounter different recommendations on how to proceed depending on the source that’s consulted (e.g., Bayesian vs frequentist inference for lab experiments, analyzing replication results.)
3. There is high degree of consensus among methodologists, but the ideas have not yet become mainstream among behavioral researchers (e.g., stimulus sampling)
4. There is high degree of consensus among methodologists but such consensus may be premature and behavioral scientists may be better off not following it (e.g., how to correct for publication bias in meta-analysis, testing for moderation in experiments, power analysis). This set will be obviously controversial.