Applying the Data Documentation Initiative (DDI) to MIDUS

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): One R03 funding option is for "development of new research technology" (National Institutes of Health Program Announcement PA-06-180). Pursuant to this option, the primary objective of the current proposal is to apply a new documentation standard called the Data Documentation Initiative (DDI) to MIDUS (Midlife in the U.S.), a national longitudinal study of over 7,000 Americans. DDI is an international effort begun in 2000 to standardize and expand the role of social science codebooks with the intent of making them available through the Internet. The DDI uses a powerful programming language called eXtensible Markup Language (XML) to create codebooks that are significant improvements over traditional text-based codebooks. In 2007, the MIDUS Administrative Core (University of Wisconsin Institute on Aging) created a DDI codebook for Project 1 of MIDUS 2 (including 2,400 variables and 4,963 participants). The specific aims of this proposal are (1) to adopt the standard for the remaining MIDUS 2 data (Projects 2-5) and bring data from the first wave of MIDUS (1995-96) into DDI compliance; (2) to enhance the scope of DDI codebooks by electronically documenting all supporting metadata where possible (creating hyperlinks to supporting files), and (3) to exploit the features of DDI in order to clarify the longitudinal nature of MIDUS by integrating Wave 1 and Wave 2 data and metadata via online analysis programs. These aims will be applied to: Project 2 (4,000 variables; 1,400 participants), which involves the daily assessments of life stressors and emotional reactions to them; Project 3 (2,400 variables; 4,500 participants), which involves the administration of a phone-based cognitive battery and an in-depth laboratory assessment of cognitive capacities; Project 4 (3,000 variables; 1,550 participants), which involves the collection of comprehensive biomarkers (neuroendocrine, cardiovascular, and immune); and Project 5 (1,000 variables; 400 participants), which involves neurological assessments (EEG, neuro-imaging). The procedures for creating DDI codebooks for the above projects as well as for MIDUS 1 include: (1) processing each Project's final dataset by a special software package called Nesstar Publisher, which creates a DDI-compliant XML file; (2) accumulating in a separate database any additional information not contained in the dataset (questionnaires, field reports, etc); (3) integrating the two sources of data (the Nesstar-derived XML file and the database of additional metadata) into a single codebook and creating a presentation which includes links to cross-referenced variables that contain skip patterns and hyperlinks to separate documentation files. These goals will be accomplished over a two-year period, following the sequential schedule for completion of data collection in Project 2 through 5. PUBLIC HEALTH RELEVANCE: The proposed project will enhance the efficient and effective public use of a national biopsychosocial study of aging. By using a powerful web-based programming language, the Data Documentation Initiative (DDI) creates interactive and comprehensive documentation that is a significant improvement over traditional text- based documentation. Adopting this technology will maximize the usefulness of the MIDUS study to researchers from diverse fields in the U.S. and beyond.