Niels Waller (2004) set out to test Meehl’s conjecture empirically. He had access to the data of more than 81,000 individuals who had completed the 567 items of the Minnesota Multiphase Personality Inventory—Revised (MMPI-2). The MMPI-2 asks people about a broad range of contents, including health, personal habits, attitudes toward sex, and extreme manifestations of psychopathology. Imagine a gender theorist who has concocted a new theory that predicts directional gender differences, that is, women will score higher on some item than men, or vice versa. Can we predict the probability of rejecting the null hypothesis in favor of the new theory? According to Meehl’s conjecture, it is about 50%. In Waller’s simulation, the computer picked the first of the 511 items of the MMPI-2 (excluding 56 for their known ability to discriminate between the sexes), determined randomly the direction of the alternative hypothesis, and computed whether the difference was significant in the predicted direction. This procedure was repeated with all 511 items. The result: 46% of the predictions were confirmed, often with very impressive p-values. Many of the item mean differences were 50–100 times larger than the associated standard errors!