Personality Testing & Discrimination Warning before Hiring: A Policy-Capturing Study
Brian Goldfeder, James Lamberti, Kevin Nolan, PhD, & Terri Shapiro, PhD Hofstra University, Department of Psychology
Introduction
Purpose of Study:
• Personality research indicates correlations between personality inventories based on the five-factor model “FFM” or “Big-Five”, and various mental disabilities (e.g., bipolar disorder).
• A major concern raised by legal scholars and I-O psychologists is the potential for disparate treatment towards individuals protected under the American with Disabilities Act (ADA).
• Recent EEOC discrimination lawsuits with brand name companies Best Buy, CVS, and Target due to misuse of personality assessments in hiring decisions.
• Therefore, the current study seeks to understand how decision-makers judgement is affected when a hypothetical warning from HR indicates that personality or interview assessments may result in people with mood disorders to be less likely hired.
Methodology
Design:
• Policy Capturing with 3 conditions: Control, Interview Warning, Personality Warning
Sample:
• Mturkers: US resident, 18 or older (n = 145)
• 57% male, avg. age = 41, 90% white, 67% 4-yr. degree
Measures:
• Person-job fit, person-organization fit, counterproductive work behaviors (CWBs), likelihood to hire, & and perceived discrimination towards oneself
Procedure:
Decision-makers rated nine hypothetical job applicant profiles based on a personality test rating, and an unstructured interview rating. Each cue was varied across three levels: low score (red), average score (yellow) or high score (green).

Results
• Linear regression analyses were conducted, generating an unstandardized regression coefficient, which was converted into a standardized coefficient, or beta weight (β). The beta weights were averaged for each condition.
• H1: One-tailed paired-samples t-tests were run. In the control condition (n = 41), interviews held a higher weight for PJ fit (t(39) = 1.13, p = .132), CWBs (t(40) = 1.48, p =.073), and likelihood to hire (t(40) = 1.63, p = .056); differences were not statistically significant. Personality held higher weight for PO fit, CWBs (t(40) = 1.03, p =.845).
• H2: One-tailed independent samples t-tests were run. For all DVs except PO fit, the beta weights were lower in the personality risk condition, but the differences were not statistically significant (PJ fit: t(94) = -0.35, p = .637; CWB: t(94) = -0.76, p = .776, likelihood to hire: t(96) = -0.74, p =. 770, PO fit: t(93) = 0.06, p = .477).
Discussion
• Warning individuals about a discrimination risk had an insignificant effect, likely due to the lack of realism for MTurkers.
• The results supported the extent research that personality inventories, a standardized method, hold less weight in decision-making than interview ratings, an unstandardized method – e.g., hiring managers believe their intuition is superior.
• The decision-maker, however, did place greater importance on personality inventory ratings for personorganization fit, indicating greater weight on personality to determine fit for company culture.

Future Research:
• Employ field studies to assess the effect of providing decision-makers with a warning about potential risks in personality testing.