NCJ Number
81211
Journal
Journal of Clinical Psychology Volume: 38 Issue: 1 Dated: (January 1982) Pages: 216-221
Date Published
1982
Length
6 pages
Annotation
Using police supervisors' ratings of 424 subordinates, this study examined the effectiveness of a linear composite of Minnesota Multiphasic Personality Inventory (MMPI) scores called the Goldberg Index in distinguishing acceptable from unacceptable police officers. The study was a preliminary step in determining valid tests for screening applicants.
Abstract
Although behavioral scientists often participate in selecting police officers, research has not proved that such involvement actually improves the selection process. Problems include the use of different criteria regarding reliability by clinicians and uncertainties about the number of individuals that comprise the baseline unacceptable officer pool. In this study, supervisors were asked to rate 424 police officers who had served from 3 to 11 years in the department. Based on this data, officers were categorized as acceptable, intermediate, and unacceptable. Supervisors' ratings were validated by information on commendations, citizens' complaints, and disciplinary actions from personnel files. Goldberg Index scores were then calculated from MMPI scores that had been part of the routine evaluation process at the police academy. Acceptables could be differentiated from unacceptables with the Goldberg Index at a high level of confidence, but validational shrinkage was observed when intermediates were added to the pool. These findings indicate that police officers are a heterogeneous population and that numerous predictive indexes are needed to screen recruits adequately. The Goldberg Index could be an excellent phase one screen which does not disqualify applicants, but identifies a pool of individuals with high potential for poor performance. Further work is necessary to determine what the Goldberg Index is measuring in a nonpsychiatric population. Tables and six references are provided.