Pratfalls in moderator research.

Considers that studies attempting to improve prediction by identifying and using moderator variables frequently incorporate methodological difficulties which could easily affect the accuracy of their conclusions. Several pitfalls are described that have appeared in published investigations: misuse of regression equation, residual scores, and validity coefficients, as well as failure to perform a complete evaluation of the moderators' actual contribution to prediction. (PsycINFO Database Record (c) 2010 APA, all rights reserved)