Should I use a repeated measures MANOVA or several repeated measures ANOVAs?

Question

I'm doing a randomised controlled trial, it has four dependant variables each measured at three time points. The data is normally distributed.

My question is how to analyse this. I'm wondering whether I should do a separate repeated measures ANOVA for each DV, or one repeated measures MANOVA? I'm using SPSS and it allows me to add multiple DVs to the repeated measures ANOVA procedure, is this the same as a MANOVA? I'm so confused..

I am aware that I could use a mixed effects model but I've chosen not to go there because I'm short on both time and knowledge!

I've read these questions but I'm still confused:

• Repeated measures ANOVA or MANOVA (SPSS)
• Should I use ANOVA or MANOVA for repeated measures experiment with two groups and several DVs?
• Repeated measures ANOVAs instead of repeated measures MANOVA w/ small N?
• Differences between MANOVA and Repeated Measures ANOVA?

Answer 1

The fact that you are asking this question probably indicates that multiple repeated measures ANOVAs are the best option. As @Pss says, MANOVAs are generally used when DVs are believed to measure the same psychological construct (e.g., depression, anxiety, extroversion), and you don't seem to believe that such a relationship exists between your 4 DVs. Further, while MANOVA is still a legitimate tool to answer research questions, it is often overused, and nowadays there are better alternatives depending on your research question. See Huang (2020) for more information on this point, and Warne (2019) for an updated general introduction to the method. Also, check out this episode of the Quantitude podcast recommended by @kjetil b halvorsen in this related post that covers the issues with MANOVA in more detail.

References

Huang, F. L. (2020). MANOVA: A procedure whose time has passed?. Gifted Child Quarterly, 64(1), 56-60.

Warne, R. (2019). A primer on multivariate analysis of variance (MANOVA) for behavioral scientists. Practical Assessment, Research, and Evaluation, 19(1).

Answer 2

Okay, I struggled to understand the same thing. The criteria for my decision was the following:

Scenarios to use MANOVAs are generally when it makes sense to "group" the dependent variables in some way. A typically example would be when a psychological scale contains constructs. There, you would typically perform a MANOVA.