Hi,

There are several possibilities and all depends on what research questions you wish to answer and to some extent the software you will use. If you have a randomized controlled trial (RCT) and you are dealing with the primary outcomes, then you should use the method that you promised to use when you registered the study. If you are doing secondary or exploratory analyses, then you may have more freedom in selecting the method. If you are interested in linear mean differences, you could probably use a linear mixed model (LMM) or a generalized estimating equations (GEE) model. In this case you should consider how to construct your variance covariance matrix for the random effects. Maybe this page can help you in this respect:

http://rpsychologist.com/r-guide-longitudinal-lme-lmer

One of the two-level models might be what you are looking for.

Regards,

Timo Törmäkangas

If your 2 dependent variables are on the same scale, you could run a repeated measures ANOVA [design: intervention_control (between) x 2_measures (within)]. If not, a MANOVA.

Furthermore, if you want to compare the 3 time points and the 2 dependent measures are on the same scale, you could run a doubly repeated measures ANOVA [design: intervention_control (between) x 3_times (within) x 2_measures (within)].

Depending on your exact situation both of the above suggestions could be helpful. This is an area of statistics called longitudinal models. I would recommend looking at the following link:

https://www.stat.ncsu.edu/people/davidian/courses/st790/notes.html

Prof Feys suggestion comes first in theses notes followed by Prof. Tormakangas suggestion. Looking here should help you sort this out. If you use anything from these Notes a thank you to Prof Davidian would be nice. Best wishes, David Booth

NICE INFORMATION THANK YOU

Hi,

the effects of two groups on two dependent variables which are measured on three time points can be determine by using Repeated measure ANOVA separately for each dependent variable.

Thanks