My question came up while preparing my data for analysis. The data come from an RCT with an intervention group and a control group. The intervention took place during 4 weeks and assessments were made at baseline, post-intervention (4-weeks) and follow-up (8-weeks). Now I have a couple of statistical outliers in some of the variables. We had several discussions on outliers in our team and I am unsure how to actually look at the outliers.

Do I now look at outliers over the entire sample or do I have to check per group? And does it make sense to look at each timepoint (baseline, post-intervention, follow-up) for outliers or over all three timepoints?

Thanks in advance, any help is very much appreciated.