How to interpret a nominal fixed effect in Linear Mixed Model analysis?
Hello there,
I am trying to interpret the estimates from a linear mixed model.
I ran an analysis where the outcome is a rating and I would like to know whether being in a group is associated with an increase or decrease of said rating.
I have a variable called group with two levels: group A and group B. It's a nominal variable with the characters as levels, so A or B.
I look at my estimates and find the following:
Question: What is group (1)? Does the rating increase by 5.240 in group A or B?
Extra Info: I looked at the variables value and label. They are both A and B, there is no numbering.
Comments
Hi jcosta,
a short answer is that the analysis uses sum contrast coding for categorical (nominal / factors, ordinal) predictors (in contrast to R which uses dummy encoding by default). We have also writen a more detail reply that hopefuly explains how to obtain results that you are interested in here: https://forum.cogsci.nl/discussion/6580/lmm-in-jasp-factor-coding
Best,
Frantisek