Temporal Stability of the Diagnostic Interview Schedule

A sample of 486 college students was administered the Diagnostic Interview Schedule (DIS) by trained lay interviewers at baseline and again approximately 9 months later. Temporal stability was assessed for 6-month, 12-month, and lifetime diagnoses of a number of disorders. As anticipated, reliability estimates tended to be lower for 12-month and 6-month diagnoses than for lifetime diagnoses. Findings indicate that the DIS is a moderately reliable instrument for assessing lifetime psychopathology over an extended time interval, although the temporal stability of lifetime diagnoses for certain disorders is probably less than is commonly assumed. To some extent, lifetime diagnostic instability appears to be attributable to "borderline" or "threshold" cases that exacerbate reliability problems in categorical diagnostic systems.