Also, the data from SMPY seems odd. I took their achievement percentages (# of individuals in top positions 40 years later), and did some rough calculations.

Also, the data from SMPY seems odd. I took their achievement percentages (# of individuals in top positions 40 years later), and did some rough calculations. The SMPY cohorts seemed to perform much better than if one had randomly selected a 'top %1' individual from the total U.S. population. So, either the longitudinal study had an effect on the cohorts, or their original sampling method was biased.