Notes
Definition and selection of comparison groups is very important. Selection bias is a danger.
Lack of randomization, manipulation, and control are all sources of weakness.
There may be another variable which is the real cause of the observed difference between groups.