Reliability and Validity of a Home-Based Self-Administered Computerized Test of Learning and Memory Using Speech Recognition.

Introduction The objective of this study is to evaluate the reliability and validity of the ReVeReTM word list recall test (RWLRT), which uses speech recognition, when administered remotely and unsupervised. Methods Prospective cohort study. Participants included 249 cognitively intact community dwelling older adults. Measures included clinician administered neuropsychological assessments at baseline and unsupervised remotely administered tests of cognition from six time-points over sixmonths. Results The RWLRT showed acceptable validity. Reliability coefficients varied across time points, with poor reliability between times 1 and 2 and fair-to-good reliability across the remaining five testing sessions. Practice effects were observed with repeated administration as expected. Discussion Unsupervised computerized tests of cognition, particularly word list learning and memory tests that use speech recognition, have significant potential for large scale early detection and long-term tracking of cognitive decline due to AD.