trying to remember my college stats class while doing some data analysis for the organization I work for. I'm trying to find out which groups of people (gender, age) our program is most effective for. We have a basic survey (Coping Self-Efficacy Scale, Chesney et al. 2006) that has three subscales. Participants took the survey pre- and post-program. Unfortunately, we don't have more follow-up data :(
Is it possible/a valid statistical procedure to run a moderation analysis on each subscale with gender or age as a moderator? I.e. Relationship between subscale1 at T1 with subscale1 at T2 as moderated by gender?
Here's the R code I'm using, where CoA is a subscale and the number is the timepoint (CoA1 or CoA2):
> uk <- read.csv( ...) > attach(uk) > summary(lm(CoA2 ~ CoA1+gender+CoA1*gender))
Thank you for the help!