All Times EDT
Keywords: mixed model, ANCOVA,rare diseases
Both mixed models and ANCOVA has been used for the analysis of repeated measures data. This talk focuses on two distinct estimands: the difference in the change from baseline in the outcome and the annualized rate of change of the outcome. Both a mixed model and an ANCOVA model can be used for the analysis of each outcome. While the mixed model makes full use of all repeated measures, there is often a desire to opt for the simplicity of the ANCOVA model in an analysis. We compare the use of the mixed model and ANCOVA in two distinct scenarios: the first is in the small sample setting with a change from baseline outcome and the second is in a moderate sample size setting with an annualized rate of change outcome. In both cases, we compare via simulation, the performance of both the mixed model and ANCOVA.