I have data collected from a pre/post intervention study with mutiple dependent variables. The depenent variables are interval type. The data was collected 'in the wild', and so there are some issues that are hampering the statistical testing I would like to conduct. The issues are:

• The number of samples in the 'pre' phase of the study is smaller that in the 'post' phase. In the pre phase, subjects were measured daily for 14 days, whilst in the post phase they were measureed for 21 days;
• Some of the daily measurements were missed or overlooked. As such, the size of the pre and post data sets is not the same accross all subjects.

Had I not encountered these issues, I would have planned to conduct a One-Way MANOVA with repeated measures to perform significance and interaction testing on the dependent variables. However, the issues outlined above make this difficult.

I have two possible solutions that I can think of:

I would love to hear your views as to whether solution 1 or 2 is better, or if there is a more effective way to deal with this situation.