Here is an interesting paper that shows that amalgamating data from subjects as is done in randomised trials can lead to severe errors in interpretation
Lack of group-to-individual generalizability is a threat to human subjects research
Aaron J. Fisher, John D. Medaglia, and Bertus F. Jeronimus
PNAS July 3, 2018. 115 (27) E6106-E6115; published ahead of print June 18, 2018. doi.org/10.1073/pnas.171197...
This is also the essential argument of our recent accepted paper.