I am pre and post testing mothers on their knowledge improvement after a safe sleep education session has been implemented. I have 8 variables I am trying to see if the answers differ. what statistical test do i use?
Hello Neco Fennoy According to the explanation you said, if you want to analyze the variables separately, then you can easily use a test such as paired-Sample T Test (if the assumptions are valid) or the nonparametric Wilcoxon test.
If you want to analyze variables as Multivariate, using a MANCOVA analysis is a good solution.
In addition to Bruce Weaver's questions, What type of measurement are you making to assess knowledge ? Are these correct/incorrect types of questions ? Or some other way to measure knowledge ?
I suggest using Cronbach's alpha on the 8 variables to see if they can be combined into a single measure (assuming the 8 variables are related). If they can be combined (alpha > 0.70), then you can use paired t test on the combined measure as recommended by others. If they are related but cannot be combined, then use repeated measures Analysis of Variance in a 2 (pre, post) X 8 (measure) design.