I have a pool of scores. In my design, two groups of subjects complete either experimental task A and a control task, or experimental task B and a control task (the same kind of control task as other group). So, Group 1 gets A and C; Group 2 gets B and C. The outcome variable scale is the same kind across tasks (0-100).
I want to compare means across all 3 tasks. I.e., do the mean scores differ significantly across tasks A, B, and C. What kind of statistical test should I use?
Issues:
- Comparing A-C or B-C would be "within-subjects", but comparing A-B is "between subjects. Comparing A-B-C involves both within and between.
- The number of scores will be larger in C than A or B, because both groups do task C. (So, if there are 50 people each in Group 1 and 2, that means 50 scores for task A, 50 for B, and 100 scores for task C.)
