The issue of whether women and minorities are attracted to S&E majors is of national interest because they now make up the majority of the labor force. Their successful completion of S&E degrees will determine the adequacy of entrants into the S&E workforce in the United States. This section reports on two longitudinal surveys of student intentions to major in S&E, by race, ethnicity, and sex. (See "Bachelor's Degrees," "Trends in Earned S&E Degrees.") The Higher Education Research Institute's (HERI) Freshman Norms Survey annually surveys a nationally representative sample of first-year students in four-year colleges and universities about their intention to major in any S&E field (HERI 1998). The National Education Longitudinal Study of 1988 (NELS:88 unpublished tabulations) tracked a large, nationally representative sample of eighth graders and identified in a follow-up survey those who were enrolled in undergraduate S&E programs (NCES 1998b).
The Freshman Norms data show that, by 1998, 47 percent of the first-year college students reporting intentions to major in S&E were women; 53 percent were men (HERI 1998). These data also show increasing racial diversity among students choosing an S&E major. By 1998, underrepresented minority groups represented 19 percent of those intending an S&E major, up from 8 percent in 1971. The trend is toward an increased percentage of black and Hispanic freshmen intending a natural science or engineering major. (See appendix table 4-9.) For example, from 1986 to 1998, the proportion of underrepresented minorities intending to major in the biological sciences rose from 10 percent of first-year college students to 18 percent. (See appendix table 4-9.)
NELS:88 corroborated the findings of the Freshman Norms Survey and showed little difference between racial and ethnic groups with regard to choosing an S&E major. NELS:88 followed students from eighth grade through high school, college, and entry into the labor force. Students who reported being enrolled in an S&E program (generally as sophomores in college) were examined to identify differences by race and sex. Between 9 and 10 percent of all racial/ethnic groups of this cohort were enrolled in S&E programs in 1994. In contrast, the study found a significant difference in the percentage of males and females enrolled in S&E programs: 12 percent of males were enrolled in such programs, whereas only 7 percent of females were enrolled in S&E fields. The gap between males and females is particularly pronounced among (or attributable mainly to) the gaps among white and Asian/Pacific Islander students; among black and Hispanic students, women are essentially on par with men. (See figure 4-7.)
Of the relatively small percentage of students who enroll as S&E majors, less than one-half complete an S&E degree within 5 years. (See "Diversity Patterns in S&E Enrollment and Degrees in the United States.") Although there may be many reasons for this, the disparity between the percentage of students who aspire to study S&E fields and the percentage who complete an undergraduate S&E degree reflects, in part, the lack of readiness of U.S. students for college-level S&E coursework.
Are U.S. freshmen ready for college-level coursework? Data from national longitudinal studies (HERI 1998; NELS:88 unpublished tabulations) and the 1994 High School Transcript Study (NCES 1997) provide some indicators of readiness: the increasing number of mathematics and science courses taken, the relatively low level of 12th-grade proficiency in science and mathematics, and the continuing need for remedial work in college.
Trend data from 1971 to 1998 on the number of high school mathematics and science courses that students have taken show that an increasing percentage of entering first-year college students have taken four years of high school mathematics and two to three years of science coursework (HERI 1998). These percentage increases have occurred across all racial groups, though they are somewhat lower among some minority groups. In 1998, between 64 and 75 percent of different subpopulations of entering first-year college students reported that they had completed four years of high school mathematicsa considerable increase from the figures reported in the previous decade. In 1984, between 37 and 65 percent of entering first-year college students reported having four years of high school math. In addition, first-year college students reported an increasing amount of high school coursework in the biological sciences. (See appendix table 4-10.) This increase in mathematics and science courses is corroborated in the 1994 High School Transcript Study, which showed that, from 1982 to 1994, rising percentages of male and female high school graduates had taken various mathematics and science courses. (See text table 4-3.)
Despite the additional mathematics and science coursetaking in high school, a relatively small percentage of 12th graders demonstrate a high level of proficiency in mathematics and science. NELS:88 tracked a representative sample of 25,000 students from eighth grade through high school, college, and entry into the labor force. This study included assessment of students' high school course-taking behavior and their mathematics and science proficiency. Analysis of the NELS data shows that in 1992 only about 37 percent of white and Asian/Pacific Islander students reached a high level of proficiency (level 4 or 5) in mathematics; an even smaller percentage of underrepresented minority students achieved this high proficiency (14 percent). In the sciences, only 25 percent of white and Asian/Pacific Islander students reached a high level (level 3) of proficiency in this exam, and even lower percentages of underrepresented minority students (8 percent) reached this level. (See appendix table 4-11.) This low level of science proficiency is also reflected in science literacy data collected with the National Science Foundation's (NSF) survey of public attitudes toward and understanding of science and technology (S&T) (see chapter 8).
Low mathematics and science proficiency is also evident among entering first-year college students. In 1997, 22 percent of first-year college students who intended an S&E major reported that they needed remedial work in mathematics; 10 percent reported they needed remedial work in the sciences. The percentages of students who need remedial work in mathematics and science have remained high over the past 20 years, with some differences by field of intended major. (See appendix table 4-12.) Students intending to major in the physical sciences and engineering report less need for remedial work. In contrast, students intending to major in the social and biological sciences, as well as in non-S&E fields, report more need for remedial work. (See figure 4-8 and appendix table 4-12.)
The American Mathematical Society's (AMS) surveys of mathematics courses (five-year incremental studies from 1970 to 1995) show an increasing percentage of remedial mathematics courses at two- and four-year colleges and a decreasing percentage of advanced-level course work at four-year institutions (NSB 1998). In the past decade, fewer students majored in mathematics (see "Bachelor's Degrees"), and universities decreased advanced-level coursework in mathematics. The forthcoming AMS survey of mathematics courses in the year 2000 should be monitored to see whether enrollment in remedial mathematics in four-year institutions continues to remain around 15 percent, or whether it decreases. (See text table 4-4.)
In contrast to intentions to major in S&E provided above, the annual fall survey of the Engineering Workforce Commission (EWC) obtains data on actual enrollment in graduate and undergraduate programs. Engineering programs generally require students to declare their major as first-year students, allowing enrollment to be used as an early indicator of undergraduate engineering degrees and interest in engineering careers.
The overall trend has been fewer students entering engineering (reflecting demographic declines in the college-age population), with a slight upturn in 1997 and 1998 (EWC 1999). At the undergraduate level, the EWC data show a declining trend in enrollment, from a high point of 441,200 students in 1983 to 356,000 students in 1996 (a 19-percent reduction). (See appendix table 4-13.) The decline was neither smooth nor continuous. Engineering enrollment stabilized for several years (1989--92) before resuming its declining trend until 1996. This declining trend turned around slightly in 1997 and 1998, with a 1.5-percent annual increase in undergraduate engineering enrollment. Trends in graduate engineering enrollment differ: graduate enrollment increased from 1979 to 1992 and then declined each year. (See figure 4-9.)
The characteristics of the community collegeflexibility, accessibility, links with industry, remediation, and low costcontribute to its broad appeal. Many students who enroll in two-year colleges are seeking certificates or associate's degrees, but some find two-year colleges an inexpensive means of completing the first two years of a college education before transferring to a four-year institution. About 22 percent of 1989/90 beginning postsecondary students who began at two-year institutions transferred to four-year institutions (NCES 1998a), thereby increasing access to higher levels of education. The majority of community colleges have links with industry; two-year engineering technology programs generally have cooperative programs with industry to train workers (Burton and Celebuski 1995). One-half of community college students are enrolled on a part-time basis. Almost all public two-year institutions provide remedial coursework, and approximately one-third of the students in these institutions are enrolled in remedial mathematics courses. (See figure 4-10.)
Since 1990, student enrollment has leveled off in all institution types that produce large numbers of S&E degrees. In contrast, over the past several decades, student enrollment in U.S. higher education institutions has increased most in two-year colleges. These institutions produce relatively few degrees, however. In 1996, 38 percent of the 15 million students in U.S. higher education were enrolled in two-year colleges, but they earned only 500,000 associate's degrees. Among beginning students at two-year colleges in the 1989/90 school year, only 24 percent had earned an associate's or higher degree by 1994 (NCES 1998a). This large disparity between the number of students enrolled and earned degrees implies high attrition rates but also highlights one of the characteristics of community collegesa large amount of coursetaking for specific skills not necessarily leading to an associate's degree. Part of this lack of persistence in completing an associate's degree is intentional; full-time students, as well as part-time, older, and night school students, may take a sequence of courses for specific skills to enter or change positions in the labor force. For a variety of other reasons, students can earn credentials below the level of associate's degree.
Among those who do earn associate-level degrees, relatively few (11 percent) earn them in S&E or engineering technology fields. In 1986--96, the number of associate's degrees in S&E fields has been modest and quite stable, ranging between 20,000 and 25,000 degrees out of approximately 450,000 to 540,000 total degrees. (See appendix table 4-16.) More numerous, however, are degrees earned in engineering technology programs (approximately 36,000 in 1996). Such engineering technology programs are mainly focused on electronics, computer technology, graphics, and mechanical engineering. (See Burton and Celebuski 1995.)
Although associate's degrees in engineering technology have been declining for about a decadereaching a low of 36,000 in 1996 (Burton and Celebuski 1995)enrollment in these programs is far higher than completed degrees would indicate. A survey of technical education in two-year colleges showed that course enrollment was about seven times higher than completed degrees (Burton and Celebuski 1995). The study also showed linkages with local industry that allow enrollees to acquire useful skills and familiarity with science, mathematics, engineering, and technology and join the industrial workforce without completing an associate's degree. Because of the importance of two-year colleges in preparing workers for high-technology employment, more needs to be known about the quality of education being provided and the attrition rates of their students.
Since the 1950s, trends in total S&E degrees earned show continual upward growth, although several fields of NS&E show a declining number of degrees in the 1990s. (See figure 4-11.) The growth occurred in two waves: the first in the 1950s and 1960s and the second in the 1990s (U.S. HEW 1956). The first growth period was the strongest; the number of degrees in S&E fields increased at an average annual rate of eight percent. Then, during the decades of the 1970s and 1980s, the total overall degrees earned fluctuated and increased at an average annual rate of less than 1 percent, followed by a second and milder growth period in the 1990s. S&E degrees at the bachelor's level increased at an average annual rate of 2.6 percent from 1990 to 1996. (See appendix table 4-17.)
The increase seen in overall S&E degrees in the past four decades actually represents divergent trends in various fields. Different fields contributed to the expansion of S&E degrees at different time periods, and several fields show a declining number of degrees in the 1990s. The number of degrees in the physical and mathematical sciences peaked in the early 1970s, slowly declined in the 1980s, and then leveled off in the 1990s. In contrast, engineering and computer science degrees peaked in the mid-1980s, quickly declined, and leveled off in the 1990s. Trends in the biological sciences showed a long, slow decline in earned degrees in the 1980s but a reversal of this trend in the 1990s. The only fields with an increasing number of earned degrees in the 1990s are psychology and the biological sciences.
The Steelman report's concern for improving the quality of undergraduate education has been of recurring national interest and has gained momentum in the past 10 years. Individual faculty, departments, professional societies, and institutions of higher education are increasingly involved in reform to enhance undergraduate teaching and the curriculum in mathematics, the various fields of sciences, engineering, and technology. Since 1992, faculties from 700 institutions of higher education have participated in one or more workshops to strengthen student interest and success in mathematics and science (Project Kaleidoscope 1999). Reforms include, for example, infusing more investigative learning into the curricula, using innovative computer laboratories and learning technologies, increasing undergraduate research experiences, and encouraging interdisciplinary collaboration in team teaching.
Reforms are directed at both science and nonscience majors. Improved introductory and advanced courses that attract and retain science majors seek both to augment the S&E workforce needed in the U.S. economy and to prepare adequate numbers of students for advanced study. Designing successful introductory courses is also aimed at strengthening the understanding of the processes and methods of science for all college students. This broader attention to curricular reform in mathematics and science courses for all students is essential for improving future K--12 teachers, public understanding of scientific issues, and citizen participation in an increasingly technological society. (See sidebar,"Institution-Wide Reform.")
Since curricular changes and facility improvements occur slowly without departmental and institutional backing, a major theme of undergraduate education reform in S&E courses is the so-called institution-wide reform. The aim of institution-wide reform is to revitalize undergraduate education on a more comprehensive, self-sustaining, and interdisciplinary basis. Recently initiated assessments of these initiatives will attempt to develop quantitative indicators on faculty, students, and institutions (Ruskus 1999). For example, faculty assessment will include the proportion of S&E faculty revising their curricula for best practices in teaching, collaborating with other faculty in developing courses, and publishing research on improved teaching and learning. Student outcomes will include the proportion of students completing S&E courses that reflect best practices, enrollment in follow-on courses, completion rates for S&E majors, and an undergraduate research experience or internship.