The interrater agreement for major diagnostic categories in studies
using DSM-I and DSM-II was usually only fair or poor. In phase one of the
DSM-III field trials the overall kappa coefficient of agreement for axis I
diagnoses of 281 adult patients was .78 for joint interviews and .66 for
diagnoses made after separate interviews; for axis II-- personality
disorders and specific developmental disorders--the coefficients of
agreement were .61 and .54. The interrater reliability of DSM--III is, in
general, higher than that previously achieved and may be due to changes in
the classification itself, the separation of axis I from axis II
conditions, the systematic description of the various disorders, and the
inclusion of diagnostic criteria.
Abstract Teaser