Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
99 APPENDIX C: Technical Notes SURVEY RESPONSE RATES* Self-Report Self-Report Year Rate Year Rate 1965 97.4 1981 95.7 1966 96.3 1982 95.3 1967 97.3 1983 95.5 1968 97.6 1984 95.1 1969 96.6 1985 94.8 1970 98.1 1986 93.5 1971 97.5 1987 93.1 1972 97.3 1988 92.9 1973 97.5 1989 92.3 1974 94.2 1990 93.6 1975 97.3 1991 94.6 1976 97.2 1992 95.1 1977 96.6 1993 94.7 1978 96.3 1994 94.6 1979 96.4 1995 94.1 1980 96.2 1996 92.8 * The rates for 1965-1995 reflect late responses. The rate for 1996 may increase slightly In We next year if additional questionnaires are received after survey closure. Self-report rates for 1980-1996 are determined Dom We "source of response" indicator In We doctorate records. Because this indicator was not coded prior to 1980, survey forms for 1965-1979 are assumed to be self-reported if "mono signed" or "marital status" is present. "Marital status" is not available Tom sources over Wan We doctorate recipient. As shown above, 92.8 percent of all doctorate recipients in 1996 completed survey forms; this percentage is referred to as the "self-report" rate. For the remaining 7.2 percent of recipients, "skeletal" forms were created with information from doctorate-granting institutions or commence- ment programs. Whether or not individuals completed the survey questionnaire, the following four data items are available for all recipients: gender, Ph.D. institution, Ph.D. field, and Ph.D. year. This report presents data obtained from all survey forms, both self-reported and skeletal. Readers should note that nonresponse in a tabulation varies according to the combination of selected variables. Higher nonresponse rates occur when any of the four variables mentioned above are cross-tabulated with another variable (e.g., educational debt) because the universe consists of the entire doctoral cohort. In other words, the 7.2 percent of Ph.D.s who did not respond to the survey are included even though their records contain minimal information. Nonresponse is generally lower when citizenship or race/ethnicity is cross-tabulated with a variable such as debt because the population is restricted to a group (e.g., U.S. citizens) that is largely drawn from self-reported forms
100 and thus more likely to have responses to the debt question. To be more precise, information on debt was not available for only 5.7 percent of U.S. citizens in ~ 996; nonresponse was low because data on both citizenship and debt were obtained mostly from self-reported forms. Nonresponse was higher for the entire ~ 996 cohort (8.9 percent) because it included the 7.2 percent of forms that were only partially filled in by institutions or staff of the National Research Council. The same was true for men (9. ~ percent) and women (8.5 percent) because gender was known even for Ph.D.s who did not complete a survey form. Cross-tabulating debt with field of doctorate would yield similarly high nonresponse rates because Ph.D. field is available for all recipients. . The percentages shown in the tables and figures in the body of this report are based only on Me number of doctorate recipients who responded to the applicable survey questions. 1 Appendix C presents nonresponse rates for the variables included in these tables and figures; it also provides descriptive explanations of the data as needed. For additional technical information, please contact: Doctorate Data Project National Opinion Research Center ~ ~ 55 East 60~ Street Chicago, ~ 60637 Phone: (773) 753-7500 Fax: (773) 753-7886 E-mail: 4800-sed~norcmail.uchicago.edu ~ Note that the percentages in Appendix Tables A-3 and A-4 are based on the total doctoral cohort because categories for "unlmown" responses are included. See the notes in front of Appendix A for further explanation of these data.
101 Baccalaureate Institutions of U.S. Minorities Table 9 is restricted to U.S. minority Ph.D.s (native and naturalized citizens) Dom 1992 to 1996 who earned baccalaureates at institutions located in the United States. Because this population constitutes only 89.0 percent of all U.S. minority Ph.D.s in this period, the totals shown in Table 9 for each group are not all inclusive. Another 9.2 percent mostly naturalized Asians and Hispanics received baccalaureates Dom foreign institutions, and the remaining I.8 percent either did not earn a baccalaureate degree or did not report this information. The totals for all U.S. minority Ph.D.s regardless of baccalaureate status are: 4,920 Asians (56.9 percent naturalized); 5,807 blacks (~.4 percent naturalized); 4,365 Hispanics (20.0 percent naturalized); and 747 American Indians (0.S percent naturalized). Country of Citizenship (for non-U.S. Ph.D.s) Country of citizenship (if missing) was first followed up in the ~ 990 survey. Consequently, nonresponse has been much lower in recent years than prior to 1990. Nonresponse was only ~ .5 percent in ~ 996, compared to 9.9 percent in ~ 989. Table ~ 3 presents data on country of citizenship. Posigraduation Plans Postgracluation status: The question on postgraduation status asks recipients to indicate whether they have made a "definite" cornrnitment, are in the process of "negotiating" with one or more organizations, or are seeking a position but have no specific prospects. Because Ph.D.s sometimes complete the survey form months ahead of graduation, it is not possible to determine the final plans of all recipients. It is quite likely that some individuals who check "negotiating" or "seeking" have obtained positions by the time of graduation. Tables 20 and 21 compare the proportion of Ph.D.s with "definite" plans and those still "seeking." Other data on postgraduation plans in this report are restncted to the group of Ph.D.s who reported "definite" plans.2 Definite commitments: Tables 22 through 27 include only those Ph.D.s who reported definite post . . grac Eaton commitments and therefore do not reflect the entire Ph.D. population. Postdoctoral location: Revisions to the survey form have resulted in significant increases in response rates for postdoctoral location during the past few years. Doctorate recipients can now check a box for "U.S." or "non-U.S." instead of providing the name and exact location of the organization with which they will be affiliated after the doctorate. This explains the much lower nonresponse since 1995 than in earlier years shown in Tables 24 through 27. See chart of item nonresponse rates for details. 2 Comparisons with the longitudinal Survey of Doctorate Recipients (SDR) show the data on "definite" postgraduation plans to be a reasonable indicator of the actual employment status of new Ph.D.s in the first year or so following receipt of the doctorate. (The SDR also conducted by the National Research Council, is a follow-up employment survey of a sample of doctorate recipients in science, engineering, and humanities fields.)
102 NONRESPONSE RATES FOR ITEMS IN TABLES Data Item Tables 1966 1971 Baccalaureate Institution (for U.S. minorities) Citizenship County of Citizenship (for non-U.S. citizens) Debt Status Doctorate Field Doctorate Institution Doctorate Year Gender Postdoctoral Location (for definite commitments) Non-U.S. citizens Early type of plans) U.S. citizens & permanent visas (employment plans) Temporary visas (employment plans) Postdoctoral Plans (e.g., definite employment vs. study) Postdoctoral Sector (for definite employment In U.S.) U.S. citizens & permanent visas Temporary visas Postdoctoral Status (e.g., definite vs. seeking) Primary Source of Graduate School Support Race/E~nicity U.S. citizens U.S. citizens & permanent visas Registered Time to Doctorate (computed) Total Time to Doctorate (computed) Table 9 Tables 7-14, 16, 17, 19, 21, 23-27 x x 2.4 1.6 Tables 12, 13 Tablesl8,19xx Tables4,5,7,8,11,15-18,20,0.00.0 22, 25, 26 x x Table 10, 14xx All tables0.00.0 Tables5,6, 16, 17, 19,21,23,270.00.0 Tables 24, 25 Tables 26, 27 Table 27 Tables 22, 23, 25-27 Tables 26, 27 Table 27 Tables 20-27 Table 17 Tables 7-10, 16, 17, 19 Tables 21,23, 27 Tables 15, 16 Tables 15, 16 xx xx xx xx xx xx xx xx xx xx x7.4 x1.7 NOTE: In 1996, 92.8 percent of new doctorate recipients completed the survey form. The item nonresponse rates in this table include the 7.2 percent of recipients who were not self-reporting. Because missing information is sometimes obtained from the doctorate-granting institutions or commencement programs, nonresponse rates for the following variables may be lower than the survey's 7.2 percent rate of nonresponse: citizenship, gender, race/ethnicity, baccalaureate institution, and total time to doctorate (derived from baccalaureate year). Field, institution, and year of doctorate are available for all recipients, as is gender. x = Year not shown in tables and figures.
103 NONRESPONSE RATES FOR ITEMS IN TABLES (Continued) . 1992 197619811986199119961996 Data Item xxxxx1.8 Baccalaureate Institution (U.S. minorities) 2.03.46.62.13.12.5 Citizenship xxx2.01.5x County of Citizenship (for non-U.S. citizens) xxxx8.8x Debt Status 0.00.00.00.00.0x Doctorate Field xxxx0.0x Doctorate Institution 0.00.00.00.00.00.0 Doctorate Year 0.00.00.00.00.0x Gender Postdoctoral Location (for definite commitments) 3.75.98.33.10.4x Non-U.S. citizens (any type of plans) 2.65.06.31.00.2x U.S. citizens & permanent visas (employment plans) 3.66.18.03.20.4x Temporary visas (employment plans) 0.80.30.50.40.2x Postdoctoral Plans (e.g., definite employment vs. study) Postdoctoral Sector (for definite employment in U.S.) 0.50.61.01.00.6x U.S. citizens & permanent visas 0.00.00.00.60.4x Temporary visas 68.19.28.59.1x Postdoctoral Status (e.g., definite vs. seeking) x x x x 12.1 x Primary Source of Graduate School Support Race/Ethnicity 4.0 4.2 1.8 1.9 1.2 1.1 U.S. citizens 4.5 4.1 2.0 2.0 1.2 x U.S. citizens & permanent visas 9.2 11.9 15.3 16.0 19.9 x Registered Time to Doctorate (computed) 1.9 3.4 7.2 4.5 5.3 x Total Time to Doctorate (computed) x = Year not shown in tables and figures.
104 Postdoctoral employment commitments in the U.S.: To be included in Tables 26 and 27, Ph.D.s must have reported definite commitments for employment. Foreign locations and employers are excluded. For temporary residents a U.S. location must have been reported. For U.S. citizens and permanent residents, unknown locations are assumed to be in the United States because of the high "stay" rates for both groups. Based on actual responses to the ~ 996 survey, 97 percent of U.S. citizens with employment or study commitments intended to remain in the United States, as did 92 percent or more of permanent residents. Primary Source of Graduate Schoo!Support In 1995 the response rate to the question on primary sources of financial support was 74.S percent. In ~ 996 the response rate jumped to 87.9 percent. This increase in response was due to a revision of the questions on sources of support. In 1995 and earlier years the questionnaire asked the respondent to identify ant! rank their sources in one question. The 1996 questionnaire asked the respondent to identify all sources of support in one question and in a separate question asked them to indicate their primary and secondary sources. The new separate question on primary/secondary sources also provided the opportunity to denote that the doctorate recipient had no primary or secondary source of support. Race/Ethr~icity Adjustments to numbers: Readers should keep in mind that fluctuations in numbers for a racial/ethnic group reflect to some degree any upward or downward change in both overall survey response and response to the racial/ethnic item. Since 1990 response to race/ethnicity has shown great improvement a result of new procedures for following up missing information. Race/ethnicity was not followed up prior to ~990. All follow-up responses received before survey closure are included in the data presented in the Summary Report for that survey. Responses arriving after closure are included in the next year's report. The extension of survey closure dates in the past four years has allowed most follow-up responses to be received in time to be included in the Summary Reports for those surveys. Postsurvey adjustments were greatest for 1990 and 1991 data, much less for 1992, and minimal for 1993. In 1994 response to the racial/ethnic item reached 97 percent by survey closure the highest rate ever. Any postsurvey adjustments for ~ 996 data will be included in next year's report, but they are expected to be very slight because of the extended closure. Updated numbers for all recent years appear in Appendix Table B-2 in this report. History of the racial/ethnic question: Although this item was first introduced to the Survey of Earned Doctorates in 1973, over 25 percent of recipients in 1973 and about 13 percent in ~ 974 either completed earlier questionnaires or provided unusable responses. Since 1975 the racial/ethnic data have been more reliable, with response rates ranging from 90.! to 97.! percent (the latter in ~ 9941. The information on race/ethnicity presented in this report is limited to the period 1977 to 1996. The racial/ethnic question has undergone several revisions over the years. in ~ 977 it was modified to correspond to a standard question format recommended by the Federal
105 Interagency Committee on Education and adopted by the Office of Management and Budget (OMB) for use in federally sponsored surveys, an explanation of the effect of these changes is detailed on page 13 of Summary Report 1977. (Note: Changes in the OMB guidelines prompted the reclassification of persons having origins in the Indian subcontinent from the white category to the Asian category.) In ~ 980 the question was further revised in two ways: (~) the Hispanic category was subdivided into Puerto Rican, Mexican American, and other Hispanic, and (2) respondents were asked to check only one racial category. (Before 1980 doctorate recipients could check more than one category to indicate their race.) The item was modified again in ~ 982 to separate the questions on race and ethnicity. Since then, respondents have been asked to first check one of the four racial group categories (American Indian, Asian, black, or white) and then indicate whether or not they are Hispanic. In this report, Ph.D.s who reported Hispanic heritage are ciassife`l as Hispanic regardless of their racial designations; the remaining Ph.D.s are then counted in the respective racial groups. (Note: Doctorate recipients who checked the category "American Indian or Alaskan Native" are identified as "American Indian" in this report.) Time to Doctorate Total time to degree (TTDJ: TTD measures the total elapsed time between the baccalaureate and the doctorate (including time not enrolled in school). TTD can be computed only for individuals whose baccalaureate year is known. Baccalaureate year is often obtained from commencement programs or doctorate institutions when not reported by the recipient. Months are now included in the computation (see note belong. Registered time to degree (RTDJ: RTD gauges the time in attendance at colleges and universities between receipt of the baccalaureate and the doctorate. Enrollment may include years of attendance not related to a recipient's doctoral program. RTD can only be computed for individuals who have provided all years of college attendance after the baccalaureate. Months are now included in the computation (see note belowJ. Note about medians: The method of computing medians has been revised. Beginning with Summary Report ~994, months (of birth, baccalaureate, and doctorates are included in the calculations whenever available; if months are missing, only years are used in the calculations. (However, medians are not computedfor years prior to 1969 because doctorate month is unavailable for aR Ph.D.s.J Medians presented in previous Summary Reports were based only on years. Some medians would be the same regardless of the method of computation, but the new method generally computes slightly different results. While differences are small (usually one- or two-tenths of ayearJ, readers should consider these differences when comparing medians presented in this report with those in earlier reports.
