Higher Education Research and Development: Fiscal Year 2010
Appendix C. Changes to Reporting Units for FY 2010 Higher Education R&D Survey
Prior to the launch of the FY 2010 Higher Education R&D (HERD) Survey, the National Science Foundation (NSF) revised the reporting procedures to require survey responses from each geographically separate institution campus headed by a president or chancellor. The goals of the change to the reporting unit were to improve the comparability of institutions appearing in the ranking tables and to facilitate peer analysis by having data available at a similar level of detail for all institutions.
Throughout the history of the survey, previously known as the Survey of R&D Expenditures at Universities and Colleges, some multicampus institutions have responded at the aggregate level rather than the campus level. As a result, the academic research and development ranking tables published annually contained a mix of institutional reporting levels. Some institutions were ranked based on their all-campus R&D total (e.g., Louisiana State University, University of Tennessee, University of Minnesota), and other institutions were ranked by their individual campus totals (e.g., University of California, University of Wisconsin, University of Massachusetts). This created an inconsistent basis for institutional ranking that prompted numerous inquiries from the public as well as complaints from institutions being ranked at the campus level. As part of its overall survey redesign project, culminating in the launch of the FY 2010 HERD Survey, NSF decided to establish and apply more consistent criteria for defining the survey's reporting unit.
After NSF sought data user and respondent input, the consensus formed that it was more accurate and desirable to have campus-level data for all institutions and to aggregate to a higher level, if needed, rather than to obtain reports at a more aggregate level. After failing to identify any existing sources with a satisfactory operational definition of "reporting unit," NSF sought to establish its own specific definition for the HERD Survey reporting unit. The requirements were that it could be verified externally and applied consistently across all U.S. institutions. Thus, the decision was made to instruct institutions to report separately for each campus headed by a president or chancellor. Branch locations with no campus president or chancellor were to be combined with the campus responsible for oversight of that branch.
Identifying Reporting Units
Prior to the FY 2010 survey, an investigation was done to identify the campuses included within all of the surveyed institutions. While building the frame for the FY 2010 HERD Survey and the FY 2010 HERD Short Form Survey (see appendix D for additional details), a list of all academic institutions in the United States and surrounding areas (Guam, Puerto Rico, and the U.S. Virgin Islands) that granted a bachelor's degree or higher was created. This list was reviewed to identify all geographically separate institution campuses, headed by a president or chancellor, that were part of an aggregated unit included in the FY 2009 Survey of R&D Expenditures at Universities and Colleges population. In addition, all institutions labeled as "all campuses" in the FY 2009 HERD Survey were investigated to see if that label was correct and to identify any campus that should report separately. Table C-1 shows all institution name changes between the FY 2009 and FY 2010 surveys. In most cases, institutions previously labeled as "all campuses" did not have more than one campus headed by a president or chancellor. Table C-1 also shows the separate campuses that were identified for 12 institutions that previously reported in the aggregate.
FY 2010 HERD Collection
Prior to fielding the FY 2010 survey, a letter announcing the upcoming launch of the new survey was sent to all institution senior administrators. The letter also announced the change in reporting level. Institutions identified by data collection staff as requiring a change to their reporting units were notified via e-mail of the change and asked to provide separate survey reports for all applicable campuses.For data users wanting to compare HERD data across university systems by aggregating member campuses, table C-2 shows all higher education R&D expenditures in the FY 2010 population, by state, institutional control, system, and institution. Additionally, table C-3 ranks institutions by total R&D expenditures using the institutional aggregations used in the previous survey's detailed statistical tables.
 After the close of FY 2010 HERD data collection, additional campuses were identified that met our criteria for separate reporting but had not been reported separately on FY 2010 survey. R&D expenditures of University of Tennessee Institute of Agriculture ($59 million) and University of Tennessee Health Science Center ($82 million) were included in the University of Tennessee Knoxville's report for FY 2010 ($292 million total). These campuses will be reported separately for FY 2011. Washington State University's campuses of Spokane, Tri-Cities, and Vancouver are each administered by chancellors but were included with Washington State University Pullman campus for FY 2010. Washington State University's accounting system is not able to identify expenditures by campus accurately, but the institution is expected to have this capability within the next few years. The respondent estimated that the vast majority of expenditures (96%) were for R&D conducted within Pullman and Spokane campuses, which share faculty and projects.