NACDA 1998 Summer Workshop

National Archive of Computerized Data on Aging (NACDA) 1998 Summer Workshop

          Using Multiple Datasets in the Study of Aging

               August 10-15, 1998 (Monday-Saturday)

           ICPSR Summer Program in Quantitative Methods
           University of Michigan, Ann Arbor, Michigan

                       Course Instructors:
      Richard T. Campbell, University of Illinois at Chicago
             John J. McArdle, University of Virginia
           Colleen A. McHorney, University of Wisconsin

In recent years the amount of data that can be brought to bear on issues of human aging has grown considerably. Specifically, comparable data are now available for selected birth cohorts that are separated by widely varying historical circumstances. Although the majority of these datasets are cross-sectional, roughly comparable longitudinal data on diverse birth cohorts are available as well. The availability of such multiple surveys involving different cohorts offers rich opportunities for cross-cohort validation of findings previously demonstrated only on single cohorts. To capitalize on the availability of these multiple datasets, sophisticated methodological and statistical techniques must be employed to address a number of thorny conceptual and technical problems, including the following: the lack of exact comparability across waves within surveys as well as across surveys themselves; relating macro-level variables that change over time to relationships at the micro level; and the comparison of differently incomplete estimates of complex models obtained from a variety of datasets. These are the broad issues that will be addressed in the workshop, which is targeted for those actively engaged in or contemplating the secondary analysis of multiple datasets.

A preliminary list of specific topics to be covered in the one-week workshop includes: (1) an overview of the available multiple datasets and a conceptual framework for ransacking them; (2) the analysis of repeated cross-sectional surveys, including traditional approaches to resolving the age-period-cohort problem and newer cohort replacement techniques; (3) multi-level models and the relationships between macro- and micro-level variables; (4) scale comparison and adjustment using item response theory; (5) configural and factorial invariance in measurement models; and (6) the analysis of incomplete data within survey waves, across survey waves within datasets, and across multiple datasets. To facilitate the learning process, the workshop will consist of a traditional didactic segment, coupled with hands-on experience using multiple datasets. The morning (3-hour) and early afternoon (1.5-hour) sessions will be devoted to lecture. The late afternoon (2-hour) sessions will involve the computer laboratory, which provides high-end Internet-linked computers reserved for workshop participants. Several examples using different multiple datasets will be keyed to each specific topic. At the end of each daily laboratory session, there will be a half-hour review and question-and-answer period.

Participants will be provided with the course outline, readings, examples, datasets, and codebooks in advance in a three-ring binder and on an accompanying CD-ROM. Multiple datasets will be selected that represent the full range of situations likely to be encountered, from the simplest case of fully replicated repeated cross-sectional surveys of the same population (e.g., pooled data from the General Social Surveys [GSS]); to the more complicated case of not fully replicated cross-sectional surveys of the same population (e.g., the Studies on Aging [SOA-I and SOA-II]); to the even more complicated case of not fully replicated longitudinal surveys of somewhat different populations (e.g., the Established Populations for the Epidemiologic Study of the Elderly [EPESE]); to the rather complex case of not fully replicated longitudinal surveys of different age cohorts (e.g., the Health and Retirement Study [HRS] and Asset and Health Dynamics Among the Oldest Old [AHEAD]).

Enrollment in the 1998 NACDA Summer Workshop is limited and competitive. Those wishing to attend should send a cover letter describing their interests in the topic, a curriculum vitae, and a completed application form. Application forms are available from the ICPSR Summer Program office. Modest stipends are available for participants.

            Please send all materials and inquiries to:

                      Henry Heitowit, Director
                        ICPSR Summer Program
                           P.O. Box 1248
                     Ann Arbor, MI, 48106-1248
                        Phone: 734-764-8392
                         Fax: 734-764-8041

               Applications are due by April 20, 1998

Funded by the National Institute on Aging, NACDA is a project of the Inter-university Consortium for Political and Social Research (ICPSR).