Online Program

Adding New Items and New Respondents: Findings from the NSF-NIH Survey of Graduate Students and Postdoctorates in Science and Engineering

Keywords: multiple respondents, survey redesign

The NSF-NIH Survey of Graduate Students and Postdoctorates in Science and Engineering (GSS) is an establishment survey that collects data at the institution-level on both graduate students and postdoctorates (postdocs). NSF wanted to expand the postdoc section of the instrument, however, exploratory research with survey respondents revealed that respondents were able to provide more accurate and complete data on graduate students compared to postdocs. This finding is consistent with research by Willimack and Nichols (2010) who argue that a significant source of measurement error in establishment surveys is due to the fact that in large companies or organizations, a single respondent may not be the most knowledgeable about all information requested in a survey. As a result, pilot research was conducted to explore the feasibility of identifying a separate or alternate respondent to provide postdoc data, and based on the results implemented the approach in the 2010 survey. This paper examines the effects on the data provided in the survey.