Feasibility of repeated administration of automated verbal‐paired‐associate memory in older adults. (31st December 2021)

• Record Type: Journal Article
• Authors: Kaula, Alexander J; Cormack, Francesca K; Taptiklis, Nick
• Is Part Of: Alzheimer's & dementia. Volume 17(2021)Supplement 6
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-12-31
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.057804
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Verbal Paired‐Associates (VPA) has traditionally been administered in the clinic by a trained rater. Automated remote administration and scoring would enable use of this cognitive task at scale, over multiple visits, to allow detection of changes in memory abilities over time. Here, we address several key aspects in a feasibility study: First, the task must have good dynamic range, with performance not at ceiling or floor in older adults. Second, in order to be sensitive to memory changes over time, the test must show good test‐retest reliability. Third, older adults should be willing and able to carry out the task and operate the technology. Method: 41 Participants aged 55‐90 years were enrolled in the study using the Prolific online study tool, agreeing to take part in a repeat‐visits study lasting five days. Five parallel sets of eight word‐pair stimuli were generated based on a model which allowed memorability characteristics to be titrated such that the task difficulty would be suitable for an older‐adult population. On each day of the study, participants were emailed a link to the day's task. We used the Neurovocalix web‐app to present word‐pairs and to automatically score their responses. A questionnaire was administered to assess the subjective experience of participants. Audio of participant responses was reviewed offline by the researcher. Result: Scores showed some improvements over time, and the range of scores conformed to our expectations based on prior modelling. Test‐retest reliability was modest between first and second sessions but became much stronger from session two onwards. Participants generally enjoyed the task and were highly motivated to perform to the best of their abilities. Questionnaire responses confirmed that the majority found the task and interaction with the automated system engaging and rewarding. Conclusion: These preliminary results suggest automatic assessment of memory works for longitudinal measurement in older adults. The change in test‐retest reliability from sessions 1‐2 to sessions 2 and onwards suggests that a familiarisation phase will be important.