Graduate Students: Some Characteristics
- Financing Graduate School
- Graduate Students' Attendance Patterns: Full- or Part-Time?
- Citizenship Issues
- Students With Disabilities
Financing Graduate School
Financial support during graduate school is often crucial. Study for graduate degrees can be expensive, and few students and/or their families can foot the bills on their own. Although this report breaks out some separate data for master's degrees, only in engineering are master's degrees sometimes terminal,  serving in the sciences mostly as way stations to the doctorate.
Students receiving U.S. doctorates support themselves and their studies through teaching and research assistantships, through "other" sources of support, and through funds of "unknown" origin. See appendix tables 4-1, 4-2, and 4-3, which report only on primary sources of maintenance (not combinations). When listing such sources, few recipients cited institutional assistantships. Students of both sexes, from all races and ethnicities, U.S. citizens or not-77 percent of those with disabilities, 69 percent of those without, 75 percent of U.S. citizens, and 71 percent of graduates from all citizenship groups-support their doctoral studies through other sources. The exception is students in the physical sciences, slightly more than half of whom are primarily supported by their work for their institutions.
About the same percentage of men and women studying for their doctorates supported themselves primarily through teaching assistantships (12 percent and 11 percent, respectively). These percentages were slightly higher for doctoral recipients in science and engineering than for recipients in all fields: 12 percent of the 17,647 men and 13 percent of the 7,537 women. A greater percentage of men than women received most of their funding through research assistantships, both in science and engineering and in other fields. (See appendix table 4-1.) Students earning nonscience and engineering doctorates (37 percent of those awarded) were less likely to be supported primarily by research assistantships (about 4 percent) than those in science and engineering (28 percent of the men and 20 percent of the women). At the high end, almost 39 percent of all engineering doctorates and 37 percent of the physical sciences doctorates were primarily supported by research assistantships.
Doctoral students' reliance on teaching and research assistantships varied according to their citizenship, racial/ethnic, and disability status. (See appendix tables 4-2 and 4-3.) In all fields, students with disabilities received financial and need-based aid about as often as did others. (See appendix table 4-4.) Just over a tenth of recipients of U.S. doctorates awarded in science and engineering are supported mainly from teaching assistantships-the low is 3 percent (agriculture) and the high, 31 percent (mathematical and computer sciences).
Some racial/ethnic groups of U.S. citizens receiving science and engineering doctorates were more likely than others to cite assistantships as their major support (appendix table 4-2). Eleven percent of whites received teaching assistantships, compared to 10 percent of American Indians, 9 percent of Hispanics, 8 percent of Asians, and 6 percent of blacks. Research assistantships, which can be an important aspect of science and engineering training, were cited as primary support by 32 percent of Asians, 23 percent of whites, 17 percent of Hispanics, 15 percent of American Indians, and 9 percent of blacks earning doctorates in those fields.
Doctorate recipients with disabilities from all citizenship groups were less likely to have received assistantships of either sort than students without disabilities. In all fields, recipients with disabilities reported smaller percentages of teaching (9 percent compared with 12 percent) and research assistantships (13 percent compared with 19 percent) than other students. The differences were similar in science and engineering fields with regard to both kinds of assistantships. Within specific science and engineering fields, however, the picture was less consistent: students with disabilities reported higher percentages of research assistantships in physical, mathematical and computer sciences, biology, and the social sciences than other students; they did similarly well in teaching assistantships in physical science and engineering. (See appendix table 4-3.)
Graduate Students' Attendance Patterns: Full- or Part-Time?
Largely because of the high cost of graduate school, many students choose to, or have no alternative but to, attend part time. Because of what Seymour and Hunter (in press) call the "disadvantage of time," the 4 percent of graduate students who have disabilities are less likely to attend graduate school full time and are more likely to attend several institutions than other students. (See SIDEBAR: Students With Disabilities Studying Science, Engineering, and Mathematics: The Time Disadvantage, and appendix table 4-4.) Although students with disabilities were slightly more likely to attend part time than others, they chose all fields at about the same rate as other students.
It is unsurprising that different kinds of doctorates take different kinds of students different amounts of time to earn. (See appendix table 4-5.) Students from all fields needed a median 10.5 years to move from their baccalaureates to their terminal degrees. These figures were less for the science and engineering fields (9.1 years) and more for all other fields (15.7 years). Although the median time between bachelor's and doctorates for women was more than for men (12.2 compared to 9.9 years for all terminal degrees and 17.0 compared to 14.2 years for nonscience and engineering fields), both sexes finished their science and engineering degrees in about 9 years. Doctoral recipients with disabilities took longer to complete their degrees than others. (See appendix table 4-32.)
Students in all fields were registered for a median 7.1 years between baccalaureate and doctoral degrees. Although the median time was less in science and engineering, some variation by field occurs. Students earned their doctorates faster in chemistry than any other science and engineering field, spending just under 6 years in study after their bachelor's. Women, spending just over 5 years, were quicker than men, who took slightly fewer than 6. 
U.S. universities occupy a position of world leadership in science and engineering doctoral education, awarding degrees to a diverse racial/ethnic group of citizens and foreign students. In 1992, whites constituted only 21 percent of the doctoral recipients who were non-U.S. citizens on temporary visas, whereas they were 88 percent of the U.S. citizens (NSF 1994, pp. 78-79). Noncitizens make up about 21 percent of the science and engineering graduate students and 33 percent of the engineers. (See appendix table 4-13.) They earned 42 percent of the doctorates in science and engineering (and 61 percent of those in engineering). (See appendix tables 4-1 and 4-2.) Data on race/ethnicity for science and engineering graduate students are available only for U.S. citizens, and-sometimes-foreign students on permanent visas; data on gender are available for all students. 
Of the total of 438,052 graduate students enrolled in science and engineering fields in 1993, 157,493 were women. (See appendix tables 4-6, 4-7 and 4-8.)
The percentage of women in these combined fields has grown steadily though slowly over the past few years, from just over 32 percent in 1988 to 36 percent in 1993. (See figure 4-3.) In science fields (excluding engineering), 44 percent of the graduate students in 1993 were women, up from 40 percent in 1988. (See figure 4-3.)
See appendix table 4-8.
Although women's representation also improved in engineering-from 13 to 15 percent-women were most outnumbered there. Women continued to dominate psychology (70 percent) and several subfields in the social and biological sciences. (See figure 4-4.)
See appendix tables 4-7 and 4-8.
Women doing graduate work in science and engineering were only slightly more likely to attend part time than men, nearly closing the gap evident in 1982, when 63 percent of women, compared to 66 percent of men, attended full time. Under a third of all students in science went part time in 1993, compared to over a third in engineering. Sixty-six percent of women and 71 percent of men attended their science graduate classes on a full-time basis. (See appendix table 4-9.) These percentages have changed very little over the last 10 years. In 1982, 62 percent of the female graduate engineering students and 60 percent of the men were enrolled full time. (See NSF 1994, p. 63.) The few graduate students studying astronomy were most likely to be enrolled full time.
Choice of Field
In 1993, female graduate students were considerably more likely to be enrolled in fields other than science and engineering than were men (Syverson and Maguire 1995, p. 4). Women were the majority in all other fields except business-ranging from 77 percent in the health fields to 55 percent in the humanities and arts.
Women's representation in science and engineering varied greatly by field. (See appendix tables 4-6, 4-7 and 4-8.) In psychology, more than two-thirds of the graduate students in 1993 were women. Women were also in the majority in biometry/epidemiology, genetics, nutrition, and several social science fields. By contrast, only 14 percent of the graduate students in physics were women.
Among the engineering fields, the highest proportion of female graduate students in 1993 was in biomedical engineering, over one-fourth. This field was followed by chemical and civil engineering, each with a female enrollment of about 20 percent; metallurgical/materials engineering and industrial engineering/management science each enrolled about 19 percent women. At the other extreme, under 10 percent of the graduate students in the mechanical and aerospace engineering fields were women.
Where They Study
Fifteen of the 20 universities enrolling the most women graduate students in science and engineering in 1993 were large state research institutions. The University of Minnesota (all campuses) enrolled the most women (2,000), followed by the University of Wisconsin (all campuses) (1,777). The private institution enrolling the most women was George Washington University (Washington, D.C.) (1,567). (See appendix table 4-11.)
Of the 332,525 U.S. citizens enrolled in graduate science and engineering programs in 1993 (both full and part time), 31,945, or 10 percent, were underrepresented minorities.  For blacks, the increase in graduate science and engineering enrollment from 1988 to 1993 was from 4 to 5 percent; for Hispanics the increase was from 3 to 4 percent; and American Indians remained under half a percentage point. Asians increased from 5 to 7 percent over those years. In 1993, whites and Asians made up 85 percent of the total enrollment. (See appendix table 4-12 and figure 4-5.)
See appendix table 4-12.
Choice of Field
The field choices of graduate students vary considerably by gender and among racial/ethnic groups. For example, 37 percent of Asian science and engineering graduate students were enrolled in engineering fields, compared with 22 percent of whites, 20 percent of Hispanics, 16 percent of American Indians, and 15 percent of blacks. (See figure 4-6.)
See appendix table 4-13
The 3,759 Asians enrolled in electrical engineering-almost 11 percent of all graduate students in this field-largely accounted for the heavy concentration of Asians in engineering.
Conversely, 37 percent of all black graduate students in science and engineering were in social science fields, compared with 30 percent of American Indians and 30 percent of Hispanics, but only 12 percent of Asians. Similarly, only 6 percent of the Asian students (and 2 percent of noncitizens) were studying psychology, whereas psychology students represented 17 percent to 22 percent of the total number of science and engineering graduate students from all other racial/ethnic groups.
Where They Study
The growing numbers  of members of minority racial/ethnic groups are differentially distributed around the country. Over 80 percent of blacks lived in metropolitan areas in 1990 (U.S. Department of Commerce 1993c), whereas nearly 9 of 10 Hispanics were concentrated in 10 states, mostly in the South and West (U.S. Department of Commerce 1993e). Asians and Pacific Islanders also live mainly in the West (U.S. Department of Commerce 1993b, 1993e); so do American Indians, more than half of whom live in six states (Oklahoma, California, Arizona, New Mexico, Alaska, and Washington) (U.S. Department of Commerce 1993d). Graduate students are also regionally concentrated. Minorities-including Asians-made up more than one-fifth of total graduate science and engineering enrollment in Mississippi, California, the District of Columbia, Georgia, and Louisiana. (See appendix tables 4-14, 4-15, 4-16, and 4-17 and NSF 1994, pp. 295-298.) Such students are enrolled in just over 80 percent of the institutions offering graduate programs, 539 out of 665. The top 10 institutions enrolled 15 percent of all minority graduate science and engineering students; the top 20 enrolled 24 percent. (See NSF 1994, p. 69.)
In 1993, three historically black colleges and universities were among the 10 institutions with the largest proportions of black science and engineering graduate students. (See appendix table 4-15 and text table 4-2.) The 10 institutions with the highest black enrollment accounted for 15 percent of all black graduate students in science and engineering fields, a proportion that has remained fairly steady for a decade.
The 25 historically black colleges and universities offering science and engineering graduate programs (4 percent of the 615 institutions offering master's in science and engineering) in 1989 awarded one in five such degrees earned by blacks. Only four historically black colleges and universities award doctoral degrees in science and engineering (Trent and Hill 1994, p. 77).
Eleven of the 50 universities that enrolled the most Hispanic graduate students were members of the Hispanic Association of Colleges and Universities. About a fifth of all Hispanic graduate students in science and engineering fields attended member institutions in both 1988 and 1993. Thirty-nine of the 50 institutions enrolling large numbers of Hispanic graduate students were located in the South, West, and Southeast, or in large urban centers such as New York or Los Angeles, where many Hispanics live. The 10 institutions with the highest Hispanic enrollment accounted for 22 percent of all Hispanic graduate students in science and engineering in the United States. Puerto Rican colleges and universities enrolled 13 percent of all Hispanic graduate students in science and engineering fields. (See appendix table 4-16.) In 1992, Puerto Rico had the highest percentage of U.S. citizen minority graduates enrolled in science and engineering, 91 percent, virtually all Hispanic (NSF 1994, pp. 295-296). 
American Indians tended to concentrate their graduate study in science and engineering in the Southwest, and a fifth attended the 10 institutions with the highest American Indian enrollment. More American Indians enroll in graduate programs in California, the state having the second largest population of American Indians in the Nation, than in any other state. (See appendix table 4-17.)
Some 13,000 Asians enrolled at the 50 universities having the most Asian graduate students in 1993. In comparison, about 15,400 underrepresented minorities attended the top 50 (in terms of concentrated enrollment) institutions for their respective groups. Seven of the top 10 universities, enrolling 3,872 of the top 10's 4,900 Asian graduate students, were in the western United States, where their population is concentrated. (See appendix table 4-14.)
Students With Disabilities
Four percent of postbaccalaureate students (including those planning master's, doctoral, and first-professional degrees) in 1993 reported a disability (Henderson 1995a, 1995b).  Proportionately, they are underrepresented compared with their 20 percent presence in the U.S. population. (See chapter 1.) Most graduate students with disabilities attended universities designed to serve all students.  Graduate students with disabilities had similar degree aspirations to others (see figure 4-7), and students with and without disabilities gravitated toward similar fields-the three most popular were education, social behavior, and business/management. (See appendix table 4-18.)
Like undergraduates with disabilities, graduates with disabilities were more likely to be veterans than other students (Henderson 1995b, p. 7). And, because incidence of disability increases with age-persons over 70 without disabilities are in a minority group (Davies 1992, analyzing Kraus and Stoddard 1989)-students with disabilities at all postsecondary levels were more likely to be older than others.
Graduate students with disabilities-like all such Americans-benefited from dramatic improvements in assistive technology. The situation for individuals with disabilities in the 1990s, in contrast to their condition only two decades ago, is vastly improved:
Then, precollege education was open to some students with disabilities but certainly not all. At the postsecondary level, individual professors and administrators were sometimes supportive, but there were no national or campus policies to make programs accessible. Science and engineering programs, with strong components of laboratory and field work, presented countless barriers (Stern and Summers 1995).
In addition, professional scientific conferences and museums were often inaccessible, and "science employment was possible but often limited" (p. vi). Assistive technology was often both ineffective and underpublicized. By now stereotypes about the limitations of persons with disabilities in science ought to disappear in the face of "closer investigation" that confirms "that pursuit of intellectual interests can surpass any limitation of physical or sensory function" (Stern and Summers 1995, p. vi). However, this is still not the case.
of racial/ethnic groups here as elsewhere in this report are limited to data
on U.S. citizens, with the exception of doctoral statistics. The latter sometimes
include foreign nationals on permanent or temporary visas.
 Better reporting of race/ethnicity, evidenced by declines in the numbers of students of "unknown race/ethnicity," could account for significant portions of the increase among underrepresented minorities. These increases may, therefore, reflect improvements in statistical quality rather than actual change.
 The 1990 census counted nearly 30 million blacks, an increase of about 4 million (1 percent) from 1980. The Hispanic population grew by 53 percent during those years. Both the numbers of Asians and Pacific Islanders and their percentage among groups in the U.S. population nearly doubled during that decade (from 3.7 to 7.3 million) (U.S. Department of Commerce 1993b, 1993c, 1993d, 1993e).
 Through a special agreement between the Puerto Rican Planning Board designated by the governor as liaison to the Census Bureau, the U.S. Census has not asked a question about race in Puerto Rico since 1950 (personal communication, Lourdes Nieves Flaim, Census Bureau, October, 1995). The 1990 Census, however, asked whether respondents could speak Spanish and English (and, if English, with how much ease or difficulty). Ninety-eight percent of respondents said they could speak Spanish; 51 percent said they didn't speak English (U.S. Department of Commerce, Bureau of the Census 1993a). Thus, Hispanic students in Puerto Rico are part of the majority ethnic culture.
 Henderson's analysis is based on data from the U.S. Department of Education, National Center for Education Statistics, National Postsecondary Student Aid Study, 1992-1993. Respondents to this telephone survey, which did not provide telecommunications devices and therefore might underreport data for individuals who are deaf or hard-of-hearing, were undergraduate, graduate, and first-professional students. Among the questions probing demographic and enrollment characteristics was one inquiring if respondents had a "functional limitation, disability, or handicap." Each survey participant answering affirmatively then faced a set of six separate questions about particular disabilities. The National Center for Education Statistics weights responses to produce national estimates for the student population. See Appendix A Technical Notes for more information.
 Most universities and colleges strive to support individuals with disabilities through special campus programs and offices. Gallaudet University (Washington, D.C.) enrolls large numbers of deaf and hard-of-hearing students in several of its graduate programs, which also admit other students; the California State University at Northridge also serves this special population within its regular curriculums. Undergraduates who are deaf and hard-of-hearing, however, can choose to enroll in a number of postsecondary institutions designed to serve them. (See chapter 3.)