Overview 
Survey Design
Survey Quality Measures
Trend Data
Availability of Data
1. Overview (FY 2007 survey cycle) 
a. Purpose
The Survey of State R&D Expenditures measures the extent of R&D activity performed and funded by each of the nation's 50 states, the District of Columbia, and Puerto Rico (collectively, state). By employing consistent, uniform definitions and collection techniques, the survey allows collection of state R&D expenditures data that are comparable nationwide, providing a source of data that were not previously widely available. Results of the survey are useful to a variety of data users interested in R&D performance, including state agencies, state policy makers, the National Science Board, the Office of Science and Technology Policy and other science policy makers, institutional researchers, and private organizations.
b. Respondents
The survey covers state government departments, agencies, independent commissions, and other entities determined to be state-run. It excludes state-run colleges and universities, which are canvassed as part of the NSF Survey of R&D Expenditures at Colleges and Universities. State-run laboratories or experiment stations controlled by state universities are excluded from the respondent universe, as are any entities determined to be nonprofit or private as defined by the U.S. Census Bureau government classification. Several industry-specific state commissions, which are generally chartered by state legislatures but are administered independently, are considered state agencies and included in the survey.
c. Key variables
- State agency or department
- State R&D expenditures
- Internal performers
- External performers
- Basic research
- Source of funds (federal, state, other)
- State R&D facilities
2. Survey Design
a. Target population and sample frame
The universe of interest covered all state departments, agencies, commissions, and dependent entities, except state-run colleges and universities, that had R&D activities for state fiscal years ending in 2007.
b. Sample design
For the FY 2006 survey, survey staff developed a list of 330 preselected agencies to be surveyed with certainty. The list was developed based on feedback during exploratory interviews conducted in FY 2006, a review of responses from previous state R&D surveys, and extensive reviews of state government Web sites. Some agencies with common missions, that are historically involved in R&D, such as departments of transportation and departments of natural resources, were selected in all states. State coordinators were then encouraged to add agencies not on the preselected list they believed were also involved with R&D. State coordinators added 103 agencies, for a total FY 2006 agency population of 433.
For the FY 2007 survey, it was decided that all 433 agencies would be included in the population. After a review of the FY 2006 responses, two duplicates were removed from the population, resulting in a total FY 2007 preselected agency count of 431. However, because the state coordinator for Maine opted to submit aggregate totals for Maine's 10 preselected agencies (as was done for the FY 2006 survey), those agencies were removed from the list, leaving a list of 421 state agencies preselected to receive the FY 2007 survey. State coordinators added 33 agencies to the respondent universe, for a total agency respondent universe of 454.
c. Data collection techniques
The survey was funded by the National Science Foundation and conducted by the U.S. Census Bureau by means of a Web-based instrument. The survey was launched in May 2008, and responses were collected through April 2009. The respondent questionnaire consisted of one screening question, intended to reduce the burden on agency respondents who did not have qualifying R&D expenditures during FY 2007, and four R&D expenditures data questions.
Before officially submitting data, state coordinators performed a final verification of aggregated agency data. All responses, including the initial agency data submissions and final state coordinator verifications, were received via the Web form or e-mail. Census staff performed basic logical edit checks and reviewed respondent comments, allowing staff to detect errors and work with state respondents to correct them.
d. Estimation techniques
All state and national totals are aggregates of reported agency data or amounts revised at the state level by state coordinators.
3. Survey Quality Measures
a. Sampling variability
Of the 444 agencies that responded to the survey, 251 (56.5%) reported having R&D activities in FY 2007. Of the 411 agencies preselected for the survey who responded, 232 (56.4%) reported having R&D. Of the 33 respondent agencies added to the survey by a state coordinator, 19 (57.6%) reported having R&D.
b. Coverage
Of the 50 states plus the District of Columbia and Puerto Rico, 46 (88.5%) provided officially verified final state data. The six entities that did not officially verify data had some or all agencies submit data. The final agency response was 444 out of 454, a response rate of 97.8%. One state, Maine, did not submit individual agency data but submitted state totals via the state coordinator.
c. Nonresponse
No statistical methods were used to account for nonresponding agencies.
d. Measurement
State coordinators in Arizona, California, Iowa, Louisiana, Tennessee, and Puerto Rico did not verify the aggregated agency data. Some reported data likely include expenditures for non-R&D activities, such as commercialization, environmental testing, or routine survey work. Some state data may also exclude minor R&D expenditure amounts from agencies not surveyed.
4. Trend Data
The FY 2007 survey was the second data collection effort using the current design. Because of differences in the survey populations, definition of covered R&D activities, and collection methods, the results of previous NSF surveys on state government R&D are not comparable with the statistics collected on the FY 2006 and FY 2007 Survey of State R&D Expenditures.
State R&D totals display considerable volatility between FY 2006 and FY 2007. While total state agency R&D expenditures increased by 20% between these two years, 28 states (and D.C. and Puerto Rico) reported overall increases, and 22 states reported overall declines. Eight states (Florida, Indiana, New Jersey, North Carolina, Ohio, Rhode Island, West Virginia, and Wyoming) reported combined agency R&D increases of more than 100% from FY 2006 to FY 2007; 5 states (Georgia, Iowa, Minnesota, New Mexico, and North Dakota) reported combined agency R&D decreases of 50% or more.
Coverage error would seem to be one of the more important factors accounting for several of the large year-to-year differences. Of the 8 states reporting at least twice the amount of R&D in FY 2007 than in FY 2006, 4 of them (Florida, Indiana, New Jersey, and Rhode Island) included agencies with large R&D expenditures in FY 2007 that had not been surveyed in FY 2006.
5. Availability of Data
a. Publications
The data from this survey are released in Detailed Statistical Tables in the series State Government R&D Expenditures, as well as in InfoBriefs.
b. Contact for more information
Additional information about this survey can be obtained by contacting:
Richard J. Bennof,
Project Officer,
Research and Development Statistics Program
Division of Science Resources Statistics
National Science Foundation
4201 Wilson Boulevard, Suite 965
Arlington, VA 22230
Phone: (703) 292-7783
E-mail: rbennof@nsf.gov
