[Computer-based assessment of memory and attention: evaluation of the memory and attention test (MAT)].

Objective A computer-based test for the assessment of working and short-term memory and selective attention was evaluated in Alzheimer patients and controls.Methods By means of the MAT, working and short-term memory for verbal, figural and episodic material are assessed through recognition tasks. Selective attention is tested through visual samples in an oddball paradigm. In 42 Alzheimer patients and 42 controls, we applied the MAT and usual reference methods for the assessment of attention and the respective memory domains.Results The correlations of the scores of the MAT memory subtests with the scores of the reference tests were moderate to high (r: 0.63-0.84). Test-retest correlations were between 0.79 and 0.90. Differentiation of Alzheimer patients and controls was good, particularly in episodic and figural short-term memory (p < 0,01). Nearly all of the probands reported that they could see and hear everything well and felt at ease during the MAT assessment.Conclusions High correlations of the MAT subtests with the reference tests indicate an acceptable construct validity. The test-retest correlations show a good reliabilty. Differentiation of Alzheimer patients and controls is fairly well. The computer-based assessment was readily accepted by the elderly probands.