I have an experiment with several factors. Factors are Condition (condition 1, condition 2), Day (1, 2), Session (1,2,3). Each subjects (N=13) undergoes 3 sessions on 2 separate days (6 sessions total). During each session, a subject undergoes condition 1 then condition 2 (or vice versa - it is counterbalanced).
My main interest is to assess if there is a difference between conditions. However, I would like to factor out possible differences between day 1 and 2, and between the different sessions (as there might be some sequence effect).
Which do you think is preferable? A repeated measures with all the factors (condition 1, condition 2), Day (1, 2), Session (1,2,3) OR collapsing sessions and days if they are not different with ANOVA/T-TEST (condition 1, condition 2). By collapsing, I mean either taking the average of the 6 measures for each subject (each of the 13 subject has 2 measures, one for condition 1 and one for condition 2), or using the values in the repeated measures ANOVA (so each of the 13 subjects has 6 sample measures for condition 1 and 6 sample measures for condition 2).