The effects of educational training are only one case of producing two groups (before and after) separated by IQ test performance on easier and harder items. The crux of the issue is whether we know that the causes between groups are genetic if they are differentiated by a Jensen effect and that the causes are environmental if they are differentiated by an anti-Jensen effect. A Jensen effect occurs when the two groups differ more as cognitive tasks become more complex. It is shown that genes vs environment must be diagnosed on a case-by-case basis, not in terms of whether a Jensen effect is present.