Laurel Brehm - Assistant Professor, UC Santa Barbara

Sep 24, 2024, 12:00 pm1:00 pm


Event Description

Your contrasts could be wrong

In the last decade, many subfields of psychology and linguistics have switched to using mixed models instead of ANOVAs to analyze their data. This is a more appropriate tool for many repeated measures study designs as it enables handling of more types of data (continuous and categorical predictors; binomial, count, and continuous dependent measures): this big switch was in general, a good thing. However, I argue that due to a misunderstanding of what statistical novices know, the problem of contrast coding-- operationalizing categorical predictors into numerical values-- was badly handled in the literature. In this talk, I present a clear demonstration of what contrast coding is, why it matters, and how to implement it in R. I also present a meta-scientific investigation into the psycholinguistic literature that shows how the majority of papers in the field fail to adequately describe their analyses.

Contributions to and/or sponsorship of any event does not constitute departmental or institutional endorsement of the specific program, speakers, or views presented.

Department of Psychology
Jason Geller