That gives enough confidence to write an entire paper about. It's ridiculous.
None of the actual data was included in the original paper. 30% of the data had to be derived because they didn't have actual IQ data for 300,000 data points. There was no statistical analysis presented. No r-sqr, no nothing.