GENLIN with binominal distrubution

#1
GENMOD with binominal distrubution

In our experimental design, we are assesing the effect of an intervention. our data is not normally distributed due to the reason that it is a mean count data. I would prefer ANCOVA with pre test scores as covariate if it was normally distributed. But I decided to use generalized linear model (GLZ) instead, and use the biniminal distribution with log link in SPSS. My question is, can I use the pretest score as covariate, post test as dependent and group as factors. As we do in GLM with ANCOVA aproach in many randomized trials in psychology.

Best
 
Last edited: