Doctoral Dissertation Research: Survey Measurement in Aging Societies

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

This research addresses the implications of an aging population for survey methodology. Psychological research into normal age-related changes in cognitive functioning has documented systematic declines across the lifespan in processing speed and working memory function. These age-related declines are likely to influence the cognitive and communicative processes involved in question answering; in fact, previous research highlighted the important role of cognitive abilities in the question-answering process. This project focuses on two specific aspects. First, psychological research indicates that older adults perform better on demanding cognitive tasks when tested in the morning rather than in the evening. If so, time of interview may interact with respondents' age in determining data quality. Secondary analyses of data from the University of Michigan Health and Retirement Study (HRS) address this question. The key prediction is that older respondents provide higher quality data when interviewed in the morning than in the evening, whereas time of day does not affect data quality for younger respondents. Second, a randomized experiment examines likely age-related differences in how respondents answer behavioral frequency questions and addresses their implications for questionnaire design. Record checks are used to asses the accuracy of respondents' answers. Controlling for other variables known to influence the selection of response strategies, the key prediction holds that older adults are less likely to employ enumeration-based response strategies than younger adults, due to the high cognitive demand associated with enumeration. Because enumeration-based strategies usually foster underreporting, whereas estimation-based strategies foster overestimating, age-related differences in strategy selection are also expected to translate into differential response biases. Specifically, older adults are expected to over report behaviors, whereas younger adults are expected to underreport behaviors relative to records, distorting actual age-related differences in behavioral frequency. Older adults are a rapidly growing segment of the population and any meaningful analysis of societal developments will need to include valid data from older adults. A solid body of research in psychology and neuroscience documents profound age-related changes in cognitive functioning and these changes present a formidable challenge for survey research: How do we adapt research strategies to ensure the collection of high quality data from older adults? Social scientists will only be able to adapt their data collection procedures to the needs of a changing population when age-related differences in respondent behavior at all steps of the research process are sufficiently well understood. This project contributes to this understanding and facilitates the development of procedures that foster high data quality, thus providing sound information about societal developments. As a Doctoral Dissertation Research Improvement award, this award also will provide support to enable a promising student to establish a strong independent research career.