Converging lines of evidence suggest that personality pathology comprises shared and unique impairments. We leveraged a large clinical sample (N = 505) and a person-centered statistical approach, ipsative change analysis, to decompose individuals’ multidimensional profiles at two time points into a metric that captures change in the elevation of the profile (i.e., impairment severity) and change in configuration of the dimensions in the profile (i.e., stylistic symptom presentation). Results demonstrated that both severity and style change were associated with overall pathology change, although the relative importance of these metrics was influenced by assessment method. Specifically, structured interview showed strong effects of severity change relative to style change, whereas self-report was less definitive. In addition, severity change was more strongly associated with change in psychosocial functioning. Results support earlier evidence of shared and unique factors in personality pathology while highlighting the influence of assessment method on models of pathology structure.