In using the SDR data set, the same doctoral field identifiers were used as those for the Doctorate Recipient File (DRF). Median salaries were determined only for a six-year cohort, 1991-1996, of doctorates that in 1997 identified themselves as being in a postdoctoral position and only for fields where a sufficient number of responses would provide reliable information. For the other figures using the SDR data, no restriction to a particular cohort was used. Data for Figure 3-2 was generated for married individuals in the year of the survey and for married individuals with children of any age under 18. Figure 1-7, containing information on the reasons why individuals were in a postdoctoral position, pertains to their current position; two categories, “additional training” and “postdoc is excepted,” were combined into a single category. Figure 2-4, which reported the employment status of 1995 postdoctoral appointees in 1997, was obtained by merging the 1995 and 1997 data to obtain responses for individuals who responded to both surveys. Figure 1-5 provides data on the median number of years individuals are in postdoctoral positions by doctoral field, and this data was available only for the 1995 survey year as a special module for the survey.
The field taxonomy for this survey is slightly different from that of the SDR or DRF since it is a survey of graduate departments and not research fields. The following departmental identifiers or a subset of those departments were used for this analysis:
Mathematics and Computer Sciences
Earth, Atmospheric, and Ocean Sciences
Physics and Astronomy
This data was extracted from the National Science Foundation's CASPER Data System. Categories were selected from menus provided by the system. Figure 1-2 contains data on all individuals holding a postdoctoral position, inde-