Measuring Cognitive Function In-Person and Remotely in Round 4 of the National Social Life, Health, and Aging Project (NSHAP).

Objectives: This paper describes the changes made to the collection of cognitive measures when the National Social Life, Health, and Aging Project (NSHAP) introduced remote modes of data collection.

Methods: In Round 4 (2021-23), the longitudinal study transitioned from being conducted in-person to collecting data via multiple modes including in-person and remote modes: web, phone, and paper-and-pencil. The team began with the measures used in Rounds 2 and 3 of NSHAP-the survey-adapted Montreal Cognitive Assessment (MoCA-SA)-and evaluated which measures could be administered remotely, introducing new measures for each cognitive subdomain, as needed, to compensate for items that could not be administered remotely.

Results: Cognitive items used in Rounds 2 and 3 that could not be administered remotely were dropped from the respective modes, and items selected from the Rush Alzheimer's Disease Center's (RADC) global cognition battery were added as substitutes. For comparison, the RADC substitute items were added to the in-person mode making it longer in R4.

Discussion: The changes in cognitive measures resulted in different numbers of cognitive items across the four modes of survey administration in Round 4. Analysts should be aware of these changes when creating a single global cognition score for the entire NSHAP sample in Round 4, and aware that there may be mode effects that could impact cognition scores.