Test-retest reliability of the dot test of visuospatial working memory in patients with schizophrenia and controls

To determine the test-retest reliability of the Dot Test of Visuospatial Working Memory, this task was administered to 29 patients with schizophrenia and 19 normal controls on two consecutive days. The test involved "copying" trials followed by "delay" recall trials. For "copying" trials, subjects saw a dot and then drew it on a blank sheet of paper. For "delay" trials, subjects drew the dot following a 10-, 20-, or 30-s delay. The distance between the stimulus and the drawn dot was measured for each trial. The composite score, termed "working memory deficit," is calculated by subtracting the average of the copying trials from the average of the delay trials. Pearson correlations revealed that overall performance in each group was comparable for days 1 and 2. Intra-class correlations of "working memory deficit" on days 1 and 2 were moderate in patients and controls, suggesting that task performance for each subject was similar on the testing days. Test-retest reliability tended to be higher for 10-s delay performance in patients and controls than for longer delay periods. Further analyses suggested that there was no significant learning effect on the task from day 1 to day 2 for either group on any measure. The Dot Test of Visuospatial Working Memory, especially the composite score, has moderate test-retest reliability and is a valuable tool that can be used to assess working memory functions in studies using a repeated-measures design.