Abstract
Despite widespread use of the Hamilton Rating Scale for Depression (HRSD; M. Hamilton, 1960), questions have been raised concerning its psychometric properties. A nonparametric item response model was used to examine how the probability of observing a specific symptom of depression changes with increases in depressive severity in a sample of depressed adults. Results showed that options from a number of items on the HRSD did not vary as a function of severity and therefore should not be viewed as indicators of depressive severity. The extent to which symptoms are expressed as a function of depressive severity carries important implications for the use of the HRSD as a measure of severity and for the debate concerning construction of depression as a continuum. Results argue against viewing depression as a simple continuum.