Abstract

Various clock scoring procedures have been developed in recent years as dementia screening measures. The current longitudinal study was developed to assess the predictive validity of the Clock Test (Tuokko, Hadjistavropoulos, Miller, Horton, & Beattie, 1995). Within a sample of subjects who initially did not meet dementia criteria, Clock Test scores (drawing, setting, and reading) distinguished between those who later met criteria for dementia as compared to subjects who did not. When impaired performance was identified as falling below cut-off on two or more subtests of the Clock Test, sensitivity and specificity were found to be 91% and 95% relative to time two diagnosis. Clock errors among the current sample were compared against normal control subjects from the Canadian Study of Health and Aging (CSHA). These comparative analyses attest to the relative normality of the clinic sample at the time of their initial assessment.