The psychologist should have recognised the inherent limitations of the particular methodology he employed and should have planned to use other selection strategies that (1) would not procedure an overestimate of validity and/or (2) would provide other kinds of evidence of the procedure’s validity. One possibility would have been to declare criterion-related validation technically infeasible due to inadequate sample sizes. Then, validation efforts could have focused on content or construct-oriented methodologies. These methods eliminate the need for cross-validation because they do not capitalize on change. The psychologist might also have employed an already-validate measure rather than, in effect, creating a new one.