I wonder if they corrected for IQ? Not just for the IQ of the subjects, more importantly the IQ of the researchers, who seem to have some basic problems with experimental design and data analysis.