Abstract
I first examined the effects of response frequency (R) on the Comprehensive System's constellation indices (Exner, 1991). R is significantly associated with 26 of the 29 constellation criteria that are based on raw numbers and significantly correlated with total scores on each of the constellations. I then examined how R impacts the external validity of the constellations. The ability of the Schizophrenia Index and the Suicide Constellation to discriminate diagnostic groups appears to be impaired when protocols deviate from average length, whereas the Hypervigilance Index (HVI) appears most diagnostic of a paranoid condition when it is relatively elevated in brief records. R also clearly mediates the relationship between the Rorschach and Minnesota Multiphasic Personality Inventory-2 (MMPI-2) criteria. In lengthy protocols, many of the cross-method convergent correlations between appropriate MMPI-2 scales and the Depression Index, Obsessive Style Index, and HVI are quite high and approach the maximum found in personality research. In brief protocols, there are strong negative correlations between these constellations and self-reports of depression and interpersonal distress. Implications from these findings for the integration of assessment methods are discussed.