Another crucial problem is that they tried to draw inferences about the causal effects of mentorship which were not justified by their data. They even acknowledged the issue, then went ahead and pretended they had identified a causal effect. And the outcome they focused on was not necessarily a good indicator of what…