Linguistics in Göttingen - A platform for empirical and theoretical linguistics

Christoph Scheepers (Glasgow): Random effects structure for confirmatory analysis: Why it's important to "keep it maximal"

I will be presenting our recent JML paper (Barr, Levy, Scheepers, & Tily, 2013; Random effects structure for confirmatory hypothesis testing: Keep it maximal), which attempts to develop standards for the use of linear mixed effects models for the purpose of analysing experimental data. The talk will include a non-technical intro to linear mixed effects models, a summary of the most important conclusions from our simulations, and a discussion of potential caveats, pitfalls, and open issues in applying mixed effects models to experimental data.