One-way anova or paired t-test on intervention study? OR BOTH?

#1
Hi all.

I'm planning a study where I get 3 separate groups of people, make them all do a satisfaction questionnaire, then give each group one different intervention (i.e. 3 different interventions total) and then get everyone to do the same satisfaction questionnaire again after the intervention.
I want to find which intervention (if any) makes the biggest improvement.

My independent variables are the type of group.
My dependent variable is the satisfaction score.

If I did a paired t-test then that I could do that 3 times, 1 t-test for each group to see if each intervention made a difference. But I read somewhere that you're not meant to do more than 2 t-tests. Is this true?

I feel like I should be measuring whether there's a difference between the groups. In this case would I do a one-way between groups ANOVA as well? I could take the mean improvement / mean change in scores (from before to after the interventions) in each group and see if there is a difference.

Is this right to do both or can I do all this in one test?

Thanks for your help; I did look through other threads but couldn't find a match to my problem.

Anna :confused:
 

Karabiner

TS Contributor
#2
Repeated measures analysis of variances, with
type of intervention as between-subjects (group)
factor, time of measurement (pre versus post) as
within-subjects factor. The interaction tells you
whether the pre-post difference is different between
groups.

With kind regards

K.