Examining 3-month test-retest reliability and reliable change using the Cambridge Neuropsychological Test Automated Battery

The Cambridge Neuropsychological Test Automated Battery (CANTAB) is a battery of computerized neuropsychological tests commonly used in Europe in neurology and psychiatry studies, including clinical trials. The purpose of this study was to investigate test-retest reliability and to develop reliable change indices and regression-based change formulas for using the CANTAB in research and practice involving repeated measurement. A sample of 75 healthy adults completed nine CANTAB tests, assessing three domains (i.e., visual learning and memory, executive function, and visual attention) twice over a 3-month period. Wilcoxon signed-rank tests showed significant practice effects for 6 of 14 outcome measures with effect sizes ranging from negligible to medium (Hedge's g: .15-.40; Cliff's delta: .09-.39). The Spatial Working Memory test, Attention Switching Task, and Rapid Visual Processing test were the only tests with scores of adequate test-retest reliability. For all outcome measures, Pearson's and Spearman's correlation coefficients ranged from .39 to .79. The measurement error surrounding difference scores was large, thus requiring large changes in performance (i.e., 1-2 SDs) in order to interpret a change score as reliable. In the regression equations, test scores from initial testing significantly predicted retest scores for all outcome measures. Age was a significant predictor in several of the equations, while education was a significant predictor in only two of the equations. The adjusted R-2 values ranged between .19 and .67. The present study provides results enabling clinicians to make probabilistic statements about change in cognitive functions based on CANTAB test performances.

Automated summary

The Cambridge Neuropsychological Test Automated Battery (CANTAB) is a commonly used computerized neuropsychological test battery in Europe. This study examined the test-retest reliability of CANTAB and developed reliable change indices and formulas for repeated measurements. The study found significant practice effects in certain aspects of cognitive function, but only a few tests showed adequate test-retest reliability. The findings suggest that large changes in performance are needed to interpret a change score as reliable.