Playing with LME 4

Lme 4 enables fitting models involving response variables that can follow a variety of distributions (e.g. binomial, Poisson). Very nice way to deal with percent correct data generated from binary responses and especially useful when ceiling or floor effects in some conditions and/or unbalanced designs appear. The afex and ez package in combination with lsmeans make the application rather straightforward (link).

21 February 2017