Integrating Information about Aging Surveys

University Of Southern California, Los Angeles CA

Investigators

Linked publications, trials & patents

Abstract

Project summary Under the current R01, we have developed the Survey Meta Data Repository (metadata.rand.org) to facilitate cross-national analyses on aging, using the family of Health and Retirement Studies around the world. Our project goals have been to complete a comprehensive digital library of survey questions and flow-charts; to provide an intelligent search engine that uses keywords, domains, and sub-domains; and to create a set of harmonized or identically defined variables. This repository also contains a library of contextual information and a Wiki system allowing researcher input. Currently, the system contains harmonized and easily searchable information on eleven health and retirement studies from 25 countries. Building on this innovative information system, our renewal application seeks to enhance the quality and usefulness of the repository by adding content and technical innovations, updating the repository with newly released data, disseminating the repository, and providing user support. Our first aim is to create harmonized variables for two newly available survey data sets, the China Health, Aging and Retirement Longitudinal Study and the Longitudinal Aging Study in India. We will also add more variables and more longitudinal data to the harmonized data sets that we created for the English Longitudinal Study of Ageing, the Survey of Health, Ageing, and Retirement in Europe, and the Korean Longitudinal Study of Aging. All our harmonized data sets are defined to have variables identical to those of the RAND HRS. Using the harmonized variables, our second aim is to provide aggregate population and sub-population estimates for quick comparisons of key indicators in an interactive, user-friendly display. Users will be able to specify level of aggregation, survey year(s), and harmonized indicator variables of interest, and our system will provide survey estimates in a table and/or graph. With this tool, users will be able to obtain aggregate-level statistical information and compare countries on key indicators without downloading micro data. Third, we aim to improve usability of the repository. We developed an intelligent search engine, enabling users to examine cross-survey concordance. We propose to further improve the presentation of cross-survey concordance information through convenient and intuitive visualization. We also aim to develop a portal on our repository that allows users to register once for access to data from multiple participating surveys. Fourth, we aim to enrich the information provided by the repository. To facilitate understanding of data sets, we will provide summary descriptions of sample designs, sample sizes, baseline response rates and retention rates for longitudinal interviews, interview modes, field periods, and other useful information such as visual aids. We have included contextual information for all survey countries in the repository and will add data and other sources of information over time. We will also create a publication archive linked to each survey, key words, domains, and sub-domains. Fifth, we will maintain and update the repository with new data and corresponding harmonized variables as they become available. This will include public release of data for recently collected surveys, currently fielded surveys, and scheduled surveys. Integrating these data involves not only importing survey meta data but also indexing all survey questions and creating flow-charts for visual illustration of question flows. Finally, we will disseminate the repository to the scientific community and provide user support. We will continue to conduct workshops at professional meetings, data workshops, and NIA population centers on how to best use the information and tools of the repository. We will also continue maintaining the help desk for users.