The Panel Study of Income Dynamics (PSID) is the world's longest running household panel survey. With over forty years of data on the same families and their descendants, the PSID is a cornerstone of the data infrastructure for empirically-based social science research in the the world. With its long-term measures of economic and social well-being, the study has allowed researchers and policy analysts to investigate the dynamism inherent in social and behavioral processes. The long panel, genealogical design, and broad content provide scientists a unique and powerful opportunity to study evolution and change within the same families over decades.

The PSID began in an era when most surveys - including the PSID - were conducted using paper and pencil; computer assisted interviewing, let alone web-based interviewing, was still decades away. A paper questionnaire was printed for each respondent, and the interviewer wrote the respondent's answers directly on the instrument. This approach was used from 1968 to 1992, when computer assisted interviewing was initiated. The completed paper interviews were returned to the University of Michigan for data entry at which time data processors would create an electronic file. For answers to questions that required coding - e.g., occupation or industry - the handwritten open-ended responses written on the instrument were not typed into the electronic data file, but instead the response was coded into a categorical variable. The instruments also contained extensive notes in the margins made by the interviewer, providing important information to the data processors about various aspects of the data being reported by the respondent that was not captured by the structured interview. These so-called "marginal notes" have been used extensively by PSID data processors to generate accurate data values. Even today, PSID data processors retrieve paper interviews from the secure facility within ISR that stores the paper interviews as they attempt to reconcile conflicting information over time.

Interview questions can often be complicated, and even potentially ambiguous to both interviewers and respondents. Because of this, a guide is prepared for interviewers that they use to help address any questions that arise when administering the survey interview to respondents. These guides, as well as other interviewer training materials, can be valuable to researchers as they interpret responses to the questions that underlie the data. The PSID has made available such information for interview waves 1997 and forward, but historical documents before this time have not been scanned/prepared and released.

The goal of this project is to preserve and enhance utilization of historical information in the PSID. Specifically, we propose to scan all completed paper questionnaires 1968 to 1992, consisting of an estimated 15 million document pages. Moreover, electronic versions of all PSID interviewer training documents, 1968 to the present, will be created and these documents will be made available to all users through the PSID website.

By converting the paper questionnaires to electronic format, the broader impacts of the project include the preservation of historically important data which is currently deteriorating, and at risk of soon being too degraded for preservation. The preservation of these data will make it easier and less expensive for PSID data processors to access this historical information, and use it to better understand and improve the quality of the data. The conversion of the interviewer materials, including training materials and "question-by-question" documents, will provide users with easy access to historically important, wave-specific information about specific questions, and data collection procedures.

A key intellectual merit is the facilitation of future scientific research - described throughout the proposal - through the ongoing improvement of data quality made possible by the preservation of these historical data, and through the resulting generation of new variables and information that would be made available to scientists.

Agency
National Science Foundation (NSF)
Institute
Division of Social and Economic Sciences (SES)
Type
Standard Grant (Standard)
Application #
1230017
Program Officer
Brian Humes
Project Start
Project End
Budget Start
2012-10-01
Budget End
2016-09-30
Support Year
Fiscal Year
2012
Total Cost
$467,697
Indirect Cost
Name
Regents of the University of Michigan - Ann Arbor
Department
Type
DUNS #
City
Ann Arbor
State
MI
Country
United States
Zip Code
48109