Background: I am currently looking to explore changes in workplace burnout profiles across two time points. In particular, I am interested to see whether change in profiles is composed of changes in level or profile or a combination of both.
To do so I have been looking at the MANOVA (and GLM) approach described by Tabachnick and Fidell. Now in essence this approach looks quite nice as the between-subjects effects give you the difference in level and the within subjects gives you the difference in profiles (flatness is no a concern in this case).
The problem is: that in their example the MANOVA approach is used to explore the difference in profiles between independent groups. In my case, however, I have longitudinal data and I am interested in the difference in level and profile between Time 1 and Time 2.
My question is: Can I still use T&Fs MANOVA approach or will the paired sample prevent this.
Advertise on Talk Stats