[em500]

Most of these studies get published based on elaborate constructions of essentially t-tests for differences in means between groups. Showing the opposite means showing no statistical difference, which is almost impossible to get published, for very human reasons.

[necovek]

My point was exactly not to do that (which is really an unsuccesfull replication), but instead to find the actual, live correlation between the same input rigourously documented and justified, and new "positive" conclusion.

As I said, harder from a research perspective, but if you can show, for instance, that sustainable companies are less profitable with a better study, you have basically contradicted the original one.