Validity and Reliability of a Pre-Employment Screening Test: The Counterproductive Behavior Index (CBI)

The Counterproductive Behavior Index (CBI) is a 120-item, true-false questionnaire developed to assess five aspects of counterproductive workplace behavior: Dependability Concerns, Aggression, Substance Abuse, Honesty Concerns, and Computer Abuse, plus an overall measure of Total Concerns. It also yields a Good Impression score. To assess predictive validity, undergraduates with significant work experience simulated persons who had each of the five counterproductive behaviors but were exercising care not to get caught trying to conceal that behavior. All differences between simulated and normative responding were highly significant, with a median sensitivity of .89 for a specificity of .90. For similar participants, construct validity correlations ranged from .37 though .72 with a median of .50, and the correlation of CBI Total Concerns with a Total Validity Index was .66. Test-retest reliabilities of the CBI scales ranged from .79 to .94 with a median correlation of .87. These compare favorably with previously reported internal consistencies (Cronbach alphas). Analysis of the CBI scores of the original normative group at different levels of Good Impression showed that none of the six Concerns scores were affected by attempts to make a good impression until the Good Impression score reached the 90th percentile.     

The impact of “non-targeted traits” on personality test faking,hiring, and workplace deviance

Evidence suggests that job applicants often “fake” on pre-employment personality tests by attempting to portray an exceedingly desirable impression in order to improve the likelihood of being selected. In the current research we shed light on the personality characteristics of those individuals who seem most likely to engage in faking. We refer to these personality variables as non-targeted traits when they are not directly targeted by the organization’s pre-employment personality test. These traits, however, may have an influence on targeted scores used for employment decision making through their effect on faking. Findings suggest that individuals will be more likely to be hired if they are low on non-targeted traits including Honesty–Humility, Integrity, and Morality, and high on Risk Taking. Such individuals also reported higher levels of workplace deviance in their current jobs. Thus, it seems that individuals low on Honesty–Humility, Integrity, and Morality, and individuals high on Risk Taking, may be most likely to engage in personality test faking, be hired, and participate in workplace deviant behaviors if these traits are not directly targeted in selection.