**Corrections**

On page 64, the formula for *d*_{D}, the 2 should be square root 2, not 2 without a square root. We thank Marc Glassman for alerting us to the error.

On page 73, the value for *r*_{D} is -.60, not .60. Note too that to compute an effect size *r* for individuals, we correlate gender with the outcome using the individual dataset to obtain a value of -0.37 (not .49). We thank Avi Kluger for alerting us to these errors.