Note on Data Sources on the S&E Labor Force
In the previous six editions of this report (formerly entitled Women and Minorities in Science and Engineering), the primary data on the S&E work force were obtained from a series of surveys conducted
by the National Science Foundation (NSF), collectively referred to as the Scientific and Technical Personnel Data System (STPDS). A careful evaluation of these surveys by the Committee on National Statistics (CNSTAT) indicated that significant
improvements were needed in the surveys in order to provide reliable estimates of the S&E population (Citro and Kalton 1989). Because of the long time required to redesign major surveys, there are no new data since the January 1992 Women and
Minorities report for two of the three surveys used in the preparation of former volumes-the Survey of Experienced Scientists and Engineers that tracked scientists and engineers identified in the 1980 Decennial Census and the Survey of Recent
Graduates in science and engineering.
The primary source of information for this chapter is the 1991 Survey of Doctorate Recipients, since the results of this survey were not available for the 1992 edition of Women and Minorities. Since the
major methodological changes recommended by CNSTAT were implemented for this survey, NSF believes that the data on the S&E doctorate population are considerably improved over prior surveys. Especially important for the purposes of this report
are significant improvements in response rates and estimation techniques that have resulted in more accurate estimates of the racial/ethnic distribution of the S&E population and a significant improvement in the questions needed to identify
individuals with disabilities. The cost of making these improvements, however, was a temporary loss in the ability to track changes in the S&E population over time. NSF is currently evaluating the feasibility of revising its estimates of
earlier Survey of Doctorate Recipients data in order to permit trend analysis.
The second primary source of information on scientists and engineers used in this chapter is the decennial census. Since the last volume of Women and Minorities was published, NSF has obtained copies
of public use tapes from the 1990 and 1980 decennial censuses. Although the definitions of some variables (including occupation and disability status) used in the decennial census are not optimal from NSF's perspective, the data file permits
estimation of statistics not previously available for the S&E population.