Paul Barrett

Real job applicants completed a 5-factor model personality measure as part of the job application process. They were rejected; 6 months later they (n 5,266) reapplied for the same job and completed the same personality measure. Results indicated that 5.2 % or fewer improved their scores on any scale on the 2nd occasion; moreover, scale scores were as likely to change in the negative direction as the positive. Only 3 applicants changed scores on all 5 scales beyond a 95 % confidence threshold. Construct validity of the personality scales remained intact across the 2 administrations, and the same structural model provided an acceptable fit to the scale score matrix on both occasions. For the small number of applicants whose scores changed beyond the standard error of measurement, the authors found the changes were systematic and predictable using measures of social skill, social desirability, and integrity. Results suggest that faking on personality measures is not a significant problem in real-world selection settings.