Comparing several means pre and post intervention - MANOVA?


I am looking to analyse some data I have collected for a piece of research. I wanted to evaluate the impact of a Mindfulness intervention upon the participants ability to be Mindful, as well as their levels of depression, anxiety and stress. My directional hypotheses were that the intervention would:

- Increase the ability to be Mindful in the participants
- Decrease stress, anxiety and depression in the participants

I collected continuous questionnaire data to measure Mindfulness, stress, anxiety and depression, both pre and post intervention with the same participants. There is no control group.

Through my reading so far, I thought that a MANOVA would be appropriate as it appears as though there are several dependent variables (depression, anxiety, stress, level of Mindfulness) and one independent variable (completing the intervention). As the same participants were used to collect the pre and post measures then it is a repeated-measures design.

I am now starting to doubt myself and would really appreciate any guidance or advice.




Well-Known Member
Manova is most commonly used when there is a suite of dependent variables that form a group where no particular variable is of special interest, for example assemblages of several species of fish at two sites. The question is do he assemblies differ?
In your case it looks as if you are interested in each DV separately. I suggest paired t tests for each DV. If you are concerned about the possibility of false positives due to multiple p values, set your significance level to something lower, say 0.05/3 or about 0.02.