How to analyze change in performance


I have a task that two groups of participants completed at Time 1 and Time 2. I was planning to conduct a repeated measures ANOVA, but also want to explore the change in performance (diff between Time 1 and Time 2) between the groups. I could conduct a t-test on this separate question, but wondering if there is another option. Any thoughts much appreciated.


Fortran must die
Why would a repeated measure ANOVA not explore the difference between time one and two? That is what repeated measure ANOVA's are used for.


Omega Contributor
When people pose this question, I always ask if group assignment was randomized or not. If not, you could have different initial values for y between groups, in which differences may mask these differences. Thus you may need to control for baseline value in your model. In addition, even if baseline values seem comparable in a non-randomized design, you could still have an omitted variable (confounder) that biases your results if not controlled for via modeling or sensitivity analysis.