Should disability items in the census be used for planning services for elders?

Purpose: The main goal of this study was to determine how well the disability questions of both the 1990 and 2000 Census correlated with a standard measure of disability. If the census questions were to correlate moderately well with a standard measure of disability, then Area Agencies on Aging (AAA) and other organizations would be able to use census information in estimating service needs for their catchment (service) area. Design and Methods: Questionnaires containing both the census disability questions and a standard measure of disability were mailed to 4,508 older adults; 1,514 completed surveys were returned. In order to assess reliability, 472 of the respondents who completed the mail survey were reinterviewed by phone. All three disability measures were collapsed into the following three categories: no needs, instrumental activities of daily living (IADL) needs only, and activities of daily living (ADL) needs. Results: All three disability measures exhibited moderate to good test-retest reliability. Using a standard measure of disability as the criterion, validity for the 1990 Census measure was quite low (Kappas of approximately 0.35). Validity for the 2000 Census measure was moderate to good (Kappas of approximately 0.60). Implications: These results suggest that the 2000 Census disability questions may be sufficiently valid for planning purposes. However, additional research with more representative samples of older adults is needed.