Racial bias in prediction with the MMPI for a juvenile delinquent population

Area of Science:

• Psychology
• Criminology
• Psychometrics

Background:

• The Minnesota Multiphasic Personality Inventory (MMPI) is a widely used psychological assessment tool.
• Concerns exist regarding potential racial bias in psychological assessments, particularly with diverse populations.
• Juvenile delinquency presents unique challenges for accurate psychological evaluation.

Purpose of the Study:

• To investigate whether the MMPI exhibits racial bias when used with male juvenile delinquents.
• To determine if MMPI scale relationships with prediction criteria differ between racial groups.
• To assess the potential adverse impact of MMPI use on Black juvenile delinquents.

Main Methods:

• Administered the MMPI to a sample of white (N=333) and Black (N=107) male juvenile delinquents.
• Collected criterion data for predictive validation.
• Developed and compared regression equations for white and Black participants.

Main Results:

• Regression analyses yielded similar prediction accuracies across racial groups.
• The weights within the regression equations showed racial differences, varying by the specific criterion.
• Constant error in prediction for racial groups decreased as criterion objectivity increased.

Conclusions:

• Unconditional statements about adverse impact of MMPI on Black delinquents are not supported.
• Evidence suggests the MMPI may not be racially biased for predictions in juvenile delinquents.
• Criterion objectivity appears to mitigate potential racial bias in MMPI predictions for this population.