Hypothetical: I have treated a sample for depression and have (continuous) depression scores at intake (Y1) and at the end of treatment (Y2). I am wondering if motivation (X) at the start of therapy (continuous) is associated with treatment outcomes or, more specifically, changes in depression (Y1 to Y2).
A common answer might be calculate difference scores between Y1 and Y2 and see if X is correlated with changes; however, it is my understanding that change scores are inherently unreliable and this might be an issue.
Moderation might be the key, but motivation moderating the relationship between time 1 depression and time 2 depression seems odd. I think I know how it would work if I had two treatment groups, but since I only have one group, I am not sure if this is applicable.
Repeated measures ANCOVA might also be another answer. Y1 and Y2 could be the 2-level within-subjet factor and X1 could be my covariate. I could compare this to a repeated measures ANOVA with simply Y1 and Y2 without the covariate.
Or is this a case for growth models or mixture modelling?
Any guidance would be appreciated!
Note: I only have SPSS, so try not to get too fancy on me!
