We examined the construction of the Yale-Brown Obsessive-Compulsive Sc
ale (Y-BOCS) and performance of subscale items based on data from 204
patients with obsessive-compulsive disorder (OCD) who participated in
a multicenter drug treatment study. Factor analysis was used to examin
e the relationship among the 10 items that make up the Y-BOCS instrume
nt. Models were computed in which both two- and three-factor solutions
were estimated. Within the two-factor solution analyses, the factor d
istribution was not so consistent as in the three-factor solution anal
yses, and a shift in the factor distribution was noted after treatment
. In the three-factor solution analyses, an independent resistance con
struct emerged in addition to the obsessive and compulsive constructs.
Examination of the posttreatment factor scores showed that the Y-BOCS
resistance items did not assess OCD symptom change as sensitively as
the rest of the Y-BOCS items did.