The proposed pilot addresses the call for projects regarding """"""""Health, Work, and Retirement."""""""" It will examine cognitive processes involved in response to survey items included in the Health and Retirement Study--Assets and Health Dynamics of the Oldest Old (HRS- AHEAD). While the importance of functional status indicators for prediction of future health change is well demonstrated, less is known about the ways older adults actually estimate their functional status, i.e., when occasional difficulty or perhaps difficulty with some component of a task leads a respondent to consider the task """"""""difficult"""""""" in a survey context. Estimation procedures may vary by task type and also according to features of respondents, such as educational attainment, race-ethnicity, and cognitive status. To examine these issues, we propose to (1) Field the ADL/IADL items included in the HRS-AHEAD in a sample of 180 respondents currently followed in a multicultural population-based sample (Northern Manhattan Aging Project); (2) Elicit concurrent opened- ended verbal protocols from respondents for each task to examine factors that may affect self-reported difficulty or need for help with functional tasks; and (3) Test a cognitive model that may account for variation in self-reported difficulty with functional tasks. We hypothesize that tasks performed irregularly (without a fixed periodicity), such as using the telephone or handling money, will entail different cognitive strategies in respondent self-reports than more regularly performed tasks (such as taking medications, dressing, or bathing). Tasks differing in the extent to which they can be modified or variably accomplished may also differ in retrieval strategies used for judgments of difficulty. This difference may be relevant for the accuracy of self-reports. To assess the latter, we will compare HRS- AHEAD survey reports of difficulty/need for help in each task to self- reported activity over the prior day, as elicited in a 24-hour time budget.
