The MIDUS-Colectica Portal

<p>DDI 3.2 is an apt metadata standard for describing related series of studies and datasets, such as those found in longitudinal research projects. MIDUS (Midlife in the U.S.), a national longitudinal study with over 12,000 cases and a broad blend of social, health, and biomarker data, has greatly benefitted from adopting DDI 3.2. In particular, this presentation describes how MIDUS has implemented DDI infrastructure to create a harmonized data extraction system that provides users of multi-disciplinary longitudinal MIDUS datasets the information needed to better analyze, interpret, and share them. Such a system allows researchers to search across datasets for variables of interest, identify and harmonize related longitudinal versions of variables, and easily create customized data extracts and codebooks that are directly related to specific research questions. One area in which DDI has proven less satisfactory is in the documentation of non-survey data captures that do not necessarily require a Question construct, per se. How are such non-survey data--including those from bio-medical, administrative, and experimental paradigms—to be described by DDI? This presentation will demonstrate some of the non-survey data and instruments in MIDUS and discuss future directions for better documenting such data using DDI