Interpretive reliability of two common MMPI-2 profiles

摘要

Users of multi-scale tests like the MMPI-2 tend not to interpret scales one at a time in a way that would correspond to standard scale-level reliability information. Instead, clinicians integrate inferences from a multitude of scales simultaneously, producing a descriptive narrative that is thought to characterize the examinee. This study was an attempt to measure the reliability of such integrated interpretations using a q-sort research methodology. Participants were 20 MMPI-2 users who responded to E-mail solicitations on professional listservs and in personal emails. Each participant interpreted one of two common MMPI-2 profiles using a q-set of 100 statements designed for MMPI-2 interpretation. To measure the “interpretive reliability” of the MMPI-2 profile interpretations, q-sort descriptions were intercorrelated. Mean pairwise interpretive reliability was .39, lower than expected, and there was no significant difference in reliability between profiles. There was also not a significant difference between within-profile and cross-profile correlations. Q-set item analysis was conducted to determine which individual statements had the most impact on interpretive reliability. Although sampling in this study was limited, implications for the field reliability of MMPI-2 interpretation are sobering.