Models for Repeated Measures

Here we have a set of data that observes the hot flash scores of patients in either the placebo group or the Gabapentin group repeatedly from times 0 (the beginning of the study) to 4 (the end of the study). Several models that are used for repeated data such as the linear mixed model and generalized estimating equations. To select the best model, I used a mixture of backward and forward model selection. Contrasts and least square means are used to find the hot flash score changes between two times. All the analyses are done on SAS.