Postdocs at Federally Funded R&D Centers: Fall 2010
Appendix A. Technical Notes
The Survey of Postdocs at Federally Funded Research and Development Centers (FFRDC Postdoc Survey) is a census of all FFRDCs in the United States that employ postdoctoral researchers (postdocs) as of 1 October of the survey year. The survey collects the total number of postdocs in each FFRDC by sex, race and ethnicity, citizenship, source of financial support, and field of research. The FFRDC Postdoc Survey is the primary source of information on postdocs employed within FFRDCs in the United States. The survey is conducted as part of the Survey on Graduate Students and Postdoctorates in Science and Engineering (GSS), which is sponsored by the National Science Foundation (NSF) and the National Institutes of Health (NIH).
The FFRDC Postdoc Survey is an establishment survey completed by FFRDC staff. The majority of respondents are administrative staff in the FFRDC director's office.
The postdoc count data are available by the following characteristics:
The survey target population includes FFRDC in the Master Government List of Federally Funded R&D Centers (see appendix B), which has been maintained by NSF since 1967. NSF maintains the master list by querying all federal agencies annually to determine changes, additions, or deletions to the list (http://www.nsf.gov/statistics/ffrdclist).
FFRDCs are engaged in basic research, applied research, development, or management of R&D activities, either upon direct request of the government or under a broad charter from the government, but in either case under the monitorship of the government. FFRDCs are operated, managed, and administered as separate organizational units within a parent organization or as separately incorporated organizations. They receive the majority of their financial support (70% or more) from the federal government, usually from one agency, and are expected to have a long-term relationship with their sponsoring agency.
FFRDCs are eligible for the FFRDC Postdoc Survey if they employed one or more postdocs as of 1 October in the survey year.
The 2010 survey universe consists of 39 FFRDCs, with 13 administered by academic institutions, 20 by nonprofit organizations, and 6 by industrial firms. Of 39 FFRDCs, 17 responded to say they had no postdocs and thus were ineligible for the 2010 survey. Twenty-two eligible FFRDCs provided the postdoc data.
The 2010 FFRDC survey collected data on the total number of postdocs employed by each center. A postdoc is defined by the GSS as an appointee who holds a PhD or equivalent doctoral degree; whose degree was awarded recently, generally within the past 5 years; whose appointment is for a limited term, generally no more than 5–7 years; who works under the supervision of a senior researcher; and whose appointment is primarily for the purpose of training in research or through scholarship.
The FFRDC survey collected data via a Web instrument. The following information was collected for all postdocs the FFRDCs employed as of 1 October 2010:
Two weeks before starting data collection on 14 February, 2011, the directors of all 39 FFRDCs were sent a letter from NSF inviting them to participate in the survey and to confirm whether the individuals listed as contact persons for previous data collections were still employed at their center or, if not, to obtain an alternative contact person's name. All survey contact materials are included in appendix C. The survey coordinators at the FFRDCs who responded to the previous survey were also sent an e-mail to verify whether they would be the respondents again for the 2010 survey. Additional follow-up contacts by e-mails and telephone calls were made to the prior-year respondents and to the FFRDC directors to identify a new contact person if the prior-year respondent was no longer available. As a result of these efforts, a contact person was identified for all 39 FFRDCs.
Access information for the 2010 Web survey instrument was e-mailed to the FFRDCs as soon as a contact person was identified for the center. During the data collection stage, 17 centers indicated that they did not have postdocs. Sixteen of these FFRDCs also reported having no postdocs in past survey cycles, and one center was new since the last survey. These FFRDCs were sent a follow-up e-mail message asking contacts to verify they do not have postdocs at their center under other job titles of "postdoctoral fellow," "postdoctoral researcher," or "research associate." All 17 centers confirmed that they had no employees meeting the definition of a postdoc working at their FFRDC. FFRDCs not employing postdocs were funded primarily by agencies within the Department of Defense and within the Department of Homeland Security (see appendix B).
The survey due date was on 14 March 2011. A series of follow-up e-mails and telephone calls was made to the nonresponding FFRDCs until the last center responded to the survey on 5 April 2011.
The survey response rate was 100%, with all 39 FFRDCs responding to the 2010 survey: 17 indicated they did not have postdocs, and 22 reported postdoc data.
Data Editing and Processing
All the data submitted by the FFRDCs were reviewed to ensure that all data fields were completed and that data were internally consistent. In cases where additional information or corrections were needed, FFRDC respondents were contacted by telephone or e-mail and asked to correct and resubmit the survey data.
Any FFRDC with cell counts that were more than 20% above or 20% below their corresponding prior year data was flagged for edit verification. Nine FFRDCs required comprehensive data review due to reported counts outside these verification limits. In all cases, FFRDC respondents confirmed their data as accurate, with changes usually explained by a significant expansion of their postdoc program from the prior year.
For each FFRDC, completed survey data were compared with the prior year's data for consistency. For any substantive disparities, respondents are sent customized e-mail messages asking them to confirm or correct the data before the final processing and tabulation of data.
Item Nonresponse and Imputation
Of 22 eligible FFRDCs, 20 reported complete data; one center only reported the total number of postdocs by citizenship and another center was not able to report data by ethnicity and race. Item nonresponse rates ranged from 0% for citizenship to 12% for ethnicity and race of the postdocs. The Web survey instrument is designed to allow respondents to indicate whether they have the details on information that allow for distinguishing missing data from zero counts. No imputation is done for missing data.
Known or Suspected Sources of Nonsampling Error
The FFRDC Postdoc Survey is a census survey and is not subject to sampling error. However, nonsampling error may arise from lack of administrative data for the categories requested on the survey and the respondents' misreporting of information. To minimize reporting errors, substantial changes in counts and inconsistent data on the completed forms are subject to callbacks to verify changes and correct anomalies in the data.
Postdoc data from the FFRDCs are available for 2005–07, 2009, and 2010. NSF releases the data from this survey through InfoBriefs and Detailed Statistical Tables in the National Center for Science and Engineering Statistics publication series Postdocs at Federally Funded R&D Centers. More detailed data for the 2010 survey and from the prior years are available upon request.
For more information about FFRDCs see NSF's Federally Funded R&D Centers Master Government List at http://www.nsf.gov/statistics/ffrdclist/. Data on Federal R&D obligations and funding to FFRDCs are available in Federal Funds for R&D report series at http://www.nsf.gov/statistics/fedfunds/. Data on R&D expenditures at FFRDCs are available at http://www.nsf.gov/statistics/ffrdc/.