O3‐03‐03: Changes to Validity of Online Cognitive Assessment in Young and Older Adults: A Comparison to Supervised Testing using the Cantab Battery

• DOI: 10.1016/j.jalz.2016.06.520
• Publication Date: 2016-10-17T07:19:24Z
• Authors: Francesca K. Cormack; Nick Taptiklis; Rosemary A. Abbott; Melis Anatürk; Ian Cartland; Lotte Coppieters; Charlotte Housden; Jennifer H. Barnett
• Keywords: Comparability; Crowdsourcing; Cognitive test

Unsupervised web-based cognitive assessment could widen and refine recruitment into clinical trials while reducing costs. However, challenges to reliability exist which are not present in supervised testing. We were interested establishing: 1) the comparability of testing 2) The development measures related participant compliance engagement. Six hundred participants between 18 70 recruited for online testing, matched 94 assessed a setting on age gender. Participants completed an adaptive test episodic memory (Paired Associates Learning - PAL) from Cantab. Demographic background data (e.g. age, education) also recorded. iPads, whereas used variety systems, including desktop computers, laptops tablet devices. There was no difference PAL errors Within there hardware platforms or browser. trial-by-trial timing showed highly variable slow reaction times number during outside bounds seen This associated with more (r=.35), younger (r=.-21). Test-retest increased when speed variability parameters those excluded. Web browser activity monitoring revealed whether tabbed different window task performance (n=200). behaviour poorer (t = -2.09, df 161.5, p-value 0.03), greater RT -3.48, 119.6, < 0.01) 3.4157, 192.9, <0.01). Behavioural metrics can be reliably identify lack engagement Good is typically older participants. A combination detailed analysis technology recommended remote Once these safeguards place, feasible produces similar results obtained settings, making this suitable function research purposes.