Choose Wisely: The Influence of Using SEM Vs. HLM on Analytic Results When Longitudinal Data Is Collected at Timepoints with Different Amounts of Occasion Variance.

I determined if results changed when using SEM vs HLM to analyze longitudinal data with increasing occasion variance. I manufactured data “collected” at four timepoints for three separate groups (N = 150; identical means; different variances). SEM/HLM produced identical coefficients; however, SEM had increasingly poor fit as variance increased