Model aligns thinking-test scores from phone and web surveys
An observational analysis of US adults around age 60 used high school and midlife information to make scores from different tasks comparable.
Limit: The model may hinder discovery of new risk factors not represented by its information
Edited by Boris Djordjevic · AgingResearch.News
US adults around age 60 who had been high school sophomores or seniors in 1980 provided data for an observational analysis of cognitive measurement. Researchers used the survey’s 2021 follow-up, with usable cognitive data from approximately 3,990 people tested only by phone, 7,000 only online and 1,670 by both methods.
Phone tasks included recalling words and repeating numbers; web tasks involved matching words or images. The researchers built a statistical model to put performance on these different tasks onto a common scale of general thinking ability. It combined task responses with information about participants from high school and midlife. Comparisons with simpler models indicated that it preserved patterns of task performance and that adding participants’ characteristics did not bias the general scores.
Why it matters
Aging surveys need ways to study cognitive function across people reached by different methods. The authors suggest their approach may be useful in other nationally representative aging studies.
Caveats
Comparability relied on assuming the same links between cognition and participant characteristics across survey modes, an assumption the checks did not contradict. The authors cautioned that the model may hinder discovery of new risk factors not represented by its existing information.
The paper
Obtaining comparable measurement of midlife cognitive functioning from disparate cognitive tasks
University of Wisconsin–Madison; Columbia University
Neuropsychology, Jun 25, 2026, Open access
Comments
Be the first to comment.