The Clock-Drawing Test (CDT) has received considerable attention from neuropsychologists in screening for dementia as a tool that has repeatedly shown good levels of discrimination between cognitively impaired and cognitively intact people (Lin et al., 2003; Royall, Mulroy, Chiodo, & Polk, 1999; Saling, Maccuspie-Moore, Anderson, & Chiu, 2002c; Shulman, 2000; Shulman, Gold, Cohen, & Zucchero, 1993; Shulman, Shedletsky, & Silver, 1986), particularly those with early-stage DAT. There are a number of approaches for administration and scoring of the CDT, and in the most widely used method, the patient receives a blank paper and is asked to draw a circle, fill in the numbers and then set up the time to indicate a specific hour, for instance "ten past eleven" (Maccuspie-Moore, 2001).
From a clinical perspective, the CDT is considered a very appealing tool. Shulman (2000), in a literature review on the qualities of the CDT, concluded that it fulfils most criteria for an 'ideal' cognitive screening tool. It has been described as a tool that can be administered quickly and easily, as one that is well tolerated by patients, is easy to score, is relatively free of the influence of language, education and culture, and enjoys acceptable levels of validity and reliability (Shulman, 2000). In a study that aimed to compare the effectiveness of three screening tools for dementia (MMSE, CASI & CDT), it was found that the Clock-Drawing Test was the most effective in detecting dementia in a low-education, non-English speaking population (Borson et al., 1999). The CDT was also found to be efficient in detecting executive dysfunction in patients' whose MMSE scores were in the normal range (Juby, Tench, & Baker, 2002). As the scoring system for the CDT takes into account qualitative aspects of performan ce (pattern of errors in addition to the total error rate), it is consistent with the notion of a "process" approach to neuropsychological evaluation, as opposed to an "achievement" approach based on right and wrong answers only (Kaplan, 1988).
Despite its sensitivity to cognitive impairments in early dementia, the ability or abilities measured by the CDT that are most responsible for its sensitivity have not been clear. While it seems to be measuring performance in a range of cognitive domains such as visuoconstructional abilities, spatial orientation, semantic memory, and executive functioning, it is usually regarded as a reliable indicator of global cognitive functioning (Shulman et al., 1993). Saling et al. (2002), in a study designed to address this question, developed several clock-related tasks which were designed to tap various cognitive components underlying clock drawing and time setting and found that the task that most discriminated between mild to moderate DAT patients and a control group was a clock-anomalies detection task. In this task, patients were shown cards with line-drawings of clock-faces that contained an incorrect feature ("Invalid clocks"). The participants were asked to point out the anomaly and DAT patients were consi derably more impaired than controls in their ability to report these anomalies (Saling, Maccuspie-Moore, Anderson, & Chiu, 2002a). According to Saling (2002) this represented an impairment in a semantic-conceptual dimension related to the graphic representation of time (Saling, Maccuspie-Moore et al., 2002a). This account is supported by other studies that have suggested that the impairment in visuoconstructive abilities seen in AD reflects an underlying impairment in conceptualization and semantic memory (Laatu, Revonsuo, Jaykka, Portin, & Rinne, 2003; Rouleau, Salmon, & Butters, 1996).