ABSTRACT
When it comes to predicting behavior at work, personality measures have complex patterns of validity results. In contrast to the highly generalizable validities of cognitive ability tests (e.g., Hunter, 1986; Salgado et al., 2003; Schmidt, 2002; Schmidt & Hunter, 2004), personality measure validities tend to be situationally specific and are influenced by a number of factors (Hough & Oswald, 2008; Tett & Christiansen, 2007). Although job complexity seems to moderate the level of the validity coefficient (Hunter & Hunter, 1984), tests of general cognitive ability tend to predict performance at a useful level across all occupations (Schmidt, 2002). In contrast, personality measures show a great deal of variability in validity coefficients across studies, and this variability is not accounted for by sampling error (Tett & Christiansen, 2007).
