The chief complaint seems to be that they only found statistical significance on one measure for one subpopulation (boys), and other measures were not significa

The chief complaint seems to be that they only found statistical significance on one measure for one subpopulation (boys), and other measures were not significant (and while the correlation for girls was not statistically significant, it was leaning the opposite direction!). Sure, maybe there's a correlation purely centered around boy's non-verbal IQ scores, but it smells a lot like a "if you look for enough correlations you'll eventually find one".