We propose to investigate the consequences of various forms of nonsampling errors on estimates of univariate, bivariate, and multivariate parameters relevant to three substantive areas: (1) functional health; (2) subjective well-being; and (3) political attitudes and behaviors. There is a long history of research by gerontologists on the first two of these three areas, and considerable research by sociologists and political scientists has focused on age differences in level and stability in political orientation and behaviors. We suspect that distributions and relationships of observed indicators of theoretically relevant concepts in each of these areas may be distorted because of measurement error; the proposed research would allow us to take account of various types of such errors, including: (1) nonresponse; (2) missing data on particular measures; and (3) inaccuracies in recorded responses. More specifically, we propose to focus on stability and change in each of the three substantive areas, on personal and situational characteristics that have effects on stability and change, and on consequences of stability and change. In these investigations we will use data that have been collected as part of the Michigan Generations Study, which includes two waves of face-to-face interviews with a representative sample of metropolitan adults plus corroborating information collected from records and other external sources, interviewers, and spouses of married respondents. We expect to address research questions that are substantively important to the understanding of stability and change over the life span and among the elderly in particular. Our primary objectives are twofold. First, we want to provide the best possible estimates of such parameters in order to advance our understanding of substantive issues. Second, we want to assess the biases that are introduced into estimates of such parameters when incorrect assumptions are made about nonsampling errors in order to alert other investigators to the importance of designing surveys in ways that reduce such errors, or at least that provide information about those errors so that their effects can be taken into account at the analysis stage.
| Herzog, A R; Wallace, R B (1997) Measures of cognitive functioning in the AHEAD Study. J Gerontol B Psychol Sci Soc Sci 52 Spec No:37-48 |
| Herzog, A R; Rodgers, W L (1989) Age differences in memory performance and memory ratings as measured in a sample survey. Psychol Aging 4:173-82 |
| Rodgers, W L; Herzog, A R; Andrews, F M (1988) Interviewing older adults: validity of self-reports of satisfaction. Psychol Aging 3:264-72 |
| Herzog, A R; Rodgers, W L (1988) Age and response rates to interview sample surveys. J Gerontol 43:S200-5 |
| Rodgers, W L; Herzog, A R (1987) Interviewing older adults: the accuracy of factual information. J Gerontol 42:387-94 |
| Herzog, A R; Dielman, L (1985) Age differences in response accuracy for factual survey questions. J Gerontol 40:350-7 |
