ABSTRACT
This chapter focuses on how to structure the random effects for a growth curve analysis. It describes how to use visual comparisons to aid in the interpretation of polynomial time terms. If the maximal model does not converge, it may be necessary to simplify the random effect structure, that is, to remove some terms from the random effects. The ranef function returns a list of two data frames, corresponding to the two sets of random effects and named by the term to the right of the pipe in the model specification. The values are the random effects, which capture how that individual’s time course differed from the group mean with respect to that particular time term. The alternative structure with Condition to the left of the pipe does not provide this kind of estimate, though an analog could be computed by averaging together the two condition-specific random effect estimates for each participant.
