CLEAN PROCESS DATA- THE LILLARD METHOD USING THE PSID

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Clean Process Data represents a specific philosophy towards the cleaning and organizing complex secondary data that leaves a fully reproducible footprint of each stage of the data preparation process as well as a final analysis product. The organizational approach underlying this process represents a significant contribution to the methodology of data preparation because the outcomes can be reproduced exactly every time. This is important because the cleaning and preparation of secondary data represents an expensive and time consuming endeavor, and this cost is increased substantially when processing steps are lost or undocumented. The presence of standardized data preparation files as part of an archived data collection not only enhances the efficiency of the research process, it also adds ongoing value to these studies as sources of secondary data. During the course of the funding period we propose to accomplish three specific tasks to illustrate the value of the Lillard approach- Initially, we propose to document fully the Lillard Clean Process Data collection as a direct archival task. The completion of the documentation for the existing Lillard PSID collection will greatly enhance the research community's ability to employ this extensive collection of harmonized longitudinal files from the PSID and help contribute to the ongoing use of this important longitudinal study. Secondly, we propose to annotate the central methodological and organizational issues operationalized by the Lillard approach and their application to econometric and social research using complex data. This task will provide a set of clear guidelines for the use of the associated data processes to aid researchers in the use of the PSID files for research. More generically, the annotated guidelines will document the value of the Lillard clean process methodologies as systematic approach with applications to the organization and management of any set of complex data. Finally, we propose to explore mechanisms under which the principles and organizational structures of Clean Process Data methodology could be applied to other secondary data collections that share a complex longitudinal structure.