Women: Their Underrepresentation and Career Differentials in Science and Engineering: Proceedings of a Conference (1987)

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.

doctoral degrees awarded, due in part to the differential time values required to complete graduate degrees and the very large number of foreign national graduate students (for the most part, men) in engineering. Table 7 (page 64) provides more up-to-date information, using l985 enrollment and B.S. degrees awarded by discipline and data on some of the smaller, nontraditional, and newer fields of engineering. All of the engineering fields are significantly below parity. Most of the larger traditional fields continue to have the lowest percentage of women: aeronautical, agricultural, civil, electrical, marine, mechan- ical, nuclear, and petroleum engineering. The exceptions are chemical, ceramic, industrial, materials, and metallurgical engineering, where participation rates for women are significantly higher but still below parity. Some of the newer emerging areas—such as bioengineering, computer engineering, environmental engineering, and systems engineer- ing—also have higher than average participation rates for women. Biological Science. As can be seen from Table 8 (page 65), women are not currently underrepresented in most of the biological science and health-related fields. Some areas (nursing, medical technology, physical therapy, and speech pathology) have always been above parity at the undergraduate level. In other areas, including biological sci- ence and pharmacy, parity has now been reached. in the hospital and health administration field, there has been more than a tenfold in- crease in the percentage of B.S. degrees awarded to women (6.7 percent in l970-7l, 75 percent in l982-83). Social Science. The picture in social science (see Table 9, page 66) is mixed. The percentage of baccalaureates awarded to women not only increased in sociology, anthropology, and psychology, but significantly exceeded parity in l983. However, the percentage of B.S. degrees awarded to women in economics, though still significantly below parity in l982-83 (about l in 3 degrees awarded), was about triple the women's participation rate of l970-7l. The women's share of bachelor's degrees in political science almost doubled between l970-7l and l982-83. Summary In science and engineering, the proportions in which undergraduate women are represented vary considerably. In most of the biological and social sciences, parity has already been reached or exceeded: projections based on college major plans of freshmen indicate that women will continue to be at least equally represented. However, in the physical sciences and engineering, women continue to be underrep- resented. Projections based on enrollments and intended majors of college-bound high school seniors and college freshmen indicate that the significant gains in participation rates made in the physical sci- ences and engineering during the l970s have not continued in the l980s, and there are some indications of regression. 59

TABLE 5: Bachelor's Awarded in Engineering, by Sex and Discipline, Year Aeronautical Chemical Civil Electrical l97l-l975 Total 8,745 l7,5l0 36,959 58,528 Women 97 487 478 563 % Women l.ll 2.78 l.29 .96 l975-76 Total l,009 3,203 8,059 9,874 Women 29 276 252 l93 % Women 2.87 8.6l 3.l2 l.95 l976-77 Total l,078 3,524 8,228 9,936 Women 28 420 429 266 % Women 2.59 ll.9l 5.2l 2.67 l977-78 Total l,l86 4,6l5 9,265 ll,2l3 Women 6l 7l6 690 435 % Women 5.l4 l5.52 7.45 3.88 l978-79 Total l,386 5,655 9,94l l2,440 Women 66 l,006 955 659 % Women 4.83 l7.79 9.67 5.30 l979-80 Total l,424 6,383 l0,442 l3,902 Women 82 l,2l5 99l 902 % Women 5.76 l9.03 9.49 6.49 l980-8l Total l,809 6,527 l0,678 l4,938 Women l29 l,252 l,l2l l,096 % Women 7.l3 l9.l8 l0.50 7.34 l98l-82 Total 2,l20 6,740 l0,524 l6,455 Women l7l l,467 l,l9l l,409 % Women 8.07 2l.77 ll.32 8.56 l982-83 Total 2,l27 7,l85 9,989 l8,049 Women l72 l,6l6 l,32l l,774 % Women 8.09 22.49 l3.22 9.83 l948-l983 TOTAL 56,l32 l3l,456 235,724 38l,9l0 Women l,044 9,050 7,846 7,944 % Women l.86 6.68 3.33 2.08 *Includes engineering technology. SOURCE: National Center for Education Statistics, Earned Degrees Con- 60

l97l-l983 Industrial Mechanical Metallurgical Mining TOTAL* l5,935 l94 l.22 40,734 295 .72 2,687 48 l.79 l,ll5 l0 .90 220,704 2,97l l.35 2,24l 87 3.88 6,84l l47 2.l4 35l 23 6.55 33l 38,774 l,3l7 3.40 8 2.4l 2,240 l43 6.38 7,703 235 350 2l 6.00 404 9 2.22 40,936 2,022 4.93 3.05 2,7l2 323 ll.9l 8,924 466 5.22 420 44 509 22 4.32 47,222 3,479 7.37 l0.48 2,804 428 l5.26 l0,l7l 603 5.93 503 8l l6.l0 600 35 5.83 53,445 4,880 9.l3 3,2l7 545 l6.94 ll,863 882 7.43 585 l00 l7.09 682 5l 7.48 69,265 6,438 9.29 3,833 756 l9.72 l3,329 l,l36 8.52 603 ll6 l9.24 750 6l 8.l3 75,000 7,699 l0.27 3,992 943 23.62 l3,922 l,220 8.76 592 98 l6.55 662 48 7.25 92,989 9,98l l0.73 3,748 987 26.33 l5,675 l,443 9.2l 645 ll7 l8.l4 597 44 7.37 89,l99 l0,95l l3.28 77,642 4,555 325,326 6,874 2.ll l6,297 707 4.34 8,398 292 l,535,088 52,974 3.45 5.87 3.48 ferred series, Washington, D.C.: U.S. Office of Education, l948-l983. 6l

TABLE 6: Percentage of Women Enrolled in Engineering, by Class, and Engineering Degrees Granted, by Degree Level, 1971-1985 Enrollments by Class Degrees Awarded Year 1st year 2nd Year 3rd Year 4th Year B.S. M.S. Ph.D. 1971 2.6 2.0 1.5 1.0 0.3 1.0 0.7 1972 3.0 2.6 2.1 1.5 1.2 1.8 0.9 1973 4.7 3.7 2.7 1.9 1.4 1.3 1.3 1974 6.7 5.4 4.0 2.8 1.8 2.5 1.1 1975 8.9 7.5 5.8 4.1 2.3 2.4 1.7 1976 10.4 •9.3 7.8 5.6 3.6 3.4 1.9 1977 11.2 10.7 9.6 7.8 4.9 3.9 2.4 1978 12.3 11.7 10.7 9.3 7.1 5.0 1.9 1979 13.5 13.2 11.8 10.5 8.9 5.5 2.2 1980 14.5 14.2 12.8 11.8 9.7 6.3 3.2 198l 15.8 15.2 13.9 12.7 10.4 6.8 3.2 1982 16.6 16.4 15.0 13.8 12.2 8.4 4.4 1983 17.0 16.9 15.4 14.5 13.2 9.0 4.7 1984 16.5 17.0 15.5 14.8 14.0 10.1 4.7 l985 l6.5 l6.7 l6.9 l4.5 l4.7 l0.2 5.7 SOURCE: Engineering Manpower Commission , Enrollment and Degree reports. Mew Yorfc: American Association of Engineering Societies, 197l-1985. TABLE 7: Percentage of Women Enrolled in Engineering Programs and B.S. Degrees Awarded to Women, 1985 Field 1st Year 4th Year Total Degrees Aeronautical 11.81 8.28 10.26 8.2 Agricultural 7.89 11.97 10.65 9.5 Architectural 18.20 20.62 19.37 15.0 aioengineering 38.34 32.78 34.39 28.7 Ceramic 26.37 26.73 27.53 23.3 Chemical 30.54 25.22 27.52 26.0 Civil 14.14 12.80 13.90 12.8 Computer 24.25 22.50 22.70 22.2 Electrical 12.60 12.32 12.76 10.8 Engineering science 16.48 16.74 17.64 18.0 Environmental 19.74 22.33 22.49 25.0 General 17.71 19.02 18.28 14.5 Industrial 30.24 27.86 30.34 28.2 Marine 11.80 9.40 9.57 6.3 Materials & Metallurgical 18.90 22.20 22.16 23.6 Mechanical 10.49 10.69 11.45 10.6 Mining 19.61 16.93 17.44 14.3 Nuclear 12.64 9.67 11.04 11.7 Other 45.00 22.67 38.91 27.3 Petroleum 14.60 11.16 12.16 11.7 Pre-engineering 17.47 18.51 17.88 __ Systems 24.14 28.59 27.30 37.6 TOTAL 16.50 14.50 16.00 14.7 SOURCE: See Table 6. 62

DI O 0 0 i i 4) a <0 CO 5 '- H 10 to u a <y .H <o o co c ^ o> 01 J3 £ .H PL, EH u CO -H (0 4J u 4J f-H • .H .H « C 01 a o> -r-l o CO 3G a 1-1 0 TJ o < 03 C >i "** O c rH i— 1 o> 10 O 6 O c O JQ J5 T3 y 01 01 o £ 4-1 1 'o cn kl c <0 .H s CO *£ u 3 m Z o> 01 01 0 o <0 6 w 10 CO £ 4-1 O H 10 0> O CT> .H 0> 10 CT< u 41 O c C rH 01 o> 0 u o u CQ 0, 00 u J ^_J CO fO *t i EH 00 OO 00 00 00 CT-, CT\ O> CT^ O> CT•. CT> vooooot*-ooo oo oo oo oo oo oo oo i i i i 0s' CO o> si Ch c o * W 01 O 41 £ CO JJ Q (0 z » c o u c u -H ft S, Q CO O <0 W S 63

TABLE 9: Sciences Percentage of B.S. Degrees Awarded to Women in Social Year Sociology Anthropology Psychology Political Economics Science l970-7l 59.3 55.5 44.6 ll.2 20.l l97l-72 57.0 54.4 46.4 ll.8 l8.8 l972-73 56.3 54.3 47.8 l3.9 l9.6 l973-74 57.3 55.5 50.5 l4.7 20.4 l974-75 58.l 57.0 52.7 l7.3 22.5 l975-76 59.3 57.4 54.3 l9.6 25.3 l976-77 60.8 58.6 56.6 22.9 28.8 l977-78 63.4 60.2 58.9 25.0 3l.l l978-79 65.2 6l.3 6l.3 27.5 33.4 l979-80 66.7 62.6 63.3 30.2 35.9 l980-8l 69.6 63.3 65.0 30.5 36.8 l98l-82 70.3 63.4 66.8 32.5 38.l l982-83 69.7 64.7 67.5 33.3 39.3 SOURCE: National Center for Education Statistics, Earned Degrees Con- ferred series, Washington, D.C.: U.S. Office of Education, l970-l983. Factors Influencing Initial Enrollments The underrepresentation of women in selected engineering and sci- ence fields has been central to a large number of studies, most of which have focused on identifying factors associated with gender dif- ferences by using aggregate data, surveys, and correlational studies. Very few experimental or quasi-experimental studies have been con- ducted. Nevertheless, important sources of information and data emerging from these studies provide a basis for understanding not only the factors related to the underrepresentation of women in science and engineering, but also methods to remedy that situation. Although there is little overt evidence of discrimination against women in the college admission process, the abundant covert explana- tions and rationalizations regarding the lower representation of women in engineering and the physical sciences usually focus on two major cognitive areas: (l) college-bound women do not have adequate prepa- ration in mathematics and physical science to pursue the rigorous demands of beginning engineering and quantitatively-oriented science programs, and (2) college-bound women score lower than men on college admissions examinations. To the extent that mathematics and physical science preparation or college admission test scores are used in the, college admissions process, these sometimes subtle requirements do significantly reduce the pool of women eligible for admission to institutions and programs that focus on engineering and science. 64

Mathematics and Physical Science Backgrounds NSF (l986b) has compiled data supporting the stronger mathematics and physical science backgrounds of high school senior and college freshman males compared to females. In addition, Surveys of High School and Beyond, l980 (U.S. Department of Education, l982) provide illustrative data, in l980, high school senior boys were more lilcely than girls to have taken courses in trigonometry (30 versus 22 per- cent), calculus (l0 versus 6 percent), chemistry (39 versus 35 percent), and physics (26 versus l4 percent). Table l0, which provides similar data for those graduating in l982, indicates that high school females continue to be less likely to attempt advanced math and physical science courses. Table ll (pages 70-7l) documents the changes that have taken place between l97l and l985 in the reported mathematics and physical science preparation of college-bound male and female seniors who took the Col- lege Entrance Examination Board exams. In the early l970s, only about one-third of the college-bound women had at least four years of high school math and two years of physical science, compared to over one-half of the college-bound males. The l985 data on college-bound students indicate a much smaller gender gap: approximately 60 percent of the college-bound women take four or more years of math and two or more years of physical science, compared to about three-fourths of the college-bound men. It should be noted that college-bound women and men reported virtually the same grades in math and physical science, both in l97l and in l985. College Board Scores Unfortunately, these changes in math and science preparation have not affected college admission test scores, which have declined in the past two decades. As shown in Table l2 (page 72), college-bound women have lower scores than men on both the SAT-Verbal and the SAT-Math. Women had slightly higher SAT-Verbal scores than men in l970 (46l versus 459) and in l97l (457 versus 454), but slightly lower scores since l972. Mean scores on the SAT-Math for college-bound women have been consistently lower than for men; 466 versus 507 in l97l, 445 versus 497 in l977, and 452 versus 499 in l985. There are differences in opinion, indeed controversy, regarding the factors accounting for the mathematical test gender differentials. The social-psychological explanation attempts to attribute the differ- ences to family socialization practices as well as temperament, in- terest, and attitudes, or to biological and experimental factors (Maccoby, l966). However, a more generally accepted sex-role social- ization analysis points to the inhibition of women to perform at high levels of accomplishment, especially in quantitative-analytical areas, ranging in source from "fear of success" stereotypes to "no women allowed or encouraged" discriminatory practices. The result is that women avoid courses and experiences at school and at home that would enhance their mathematical and science backgrounds and compe- tencies—the differential coursework hypothesis. 65

0 M <_^ •C M o .rH CO CO 1 CO 3 CJ VO vV rH •H 00 ^4 3 CO f^. (^ Q\ fel U • • • £; H CTi CM t^ r? tH CM O CO 5 10 • • • K _£ CJ vo co in ^* 8- oi CO CO e . o "i^ "^ •H CO Oi § W CTl iH VO i JC CO £ VO CM rH c o S rH CM iH "r« CO rH en 10 CT) CTi OO c en 1 OO CTi O M •D re M C 10 0 00 . kl CD CJ JJ r-4 kl n •H 0) ^1 0) E f CM VO .H _Q E £ • • • O u ^" in «n CO c Tl o <J> 00 0 c d> 0< 4 Jj •H • • • •iH OI kl CM in o in EH CM CM CM ^ co rH kl o> , JJ •H • JJ co *j Q, •* E in ^ m -JJ % co IH W • • • o «O $ M J3 in vo •»»• c 00 O CO CO CO 9t kl 10 SE iH ki ; 8 0) Cn n CM m ^" in CO •o tk U SN <0 •*4 0) <c ro m ro 0 , O *!• VO c MH 0 rH (4-) c O • • • g o 0) •H CO VO Cft •H ea rH rH rH 0 e. s >, & c jj ^ "^ ^ JJ t3 *. "^ a> Q) >i o '"^ *" CD c E CM 9\ •& CJ en M O • • • c o DJ CO kl O O> cj I in in in 10 iH > 0 00 0 0 ll •a -H co r^ o •H Ql < m r- r- oo o ^ JJ §c CO r ^ •H u k.1 o> — •H i— i Oi JJ i-l 0) > JJ X <0 •0 a> 10 0i CJ O <o o> e kl u u u c o £ CO 0) X! c •H C 0) r- iH ro o> co a> ao m i-i •H • >*-i ^ S- JJ • • • -H |>i -«-i • • i> U CO ° 10 rH l~- VO CT, O £ O r^ o m co • CO CM £ vo vo vo CO & co vo r^ vo rH D o> oo C J EH ""' -^ Q C \ 3 a «J •.H rH O 3 C 0 "° 10 M V 0) o 1"H JJ •H O 0) 10 iH o> <o rH JJ 0 rH e -H 6 SOURCE Washin 10 W 3 J TJ 00 10 10 U J <0 vV >J 10 0) « -H \ C g ,; „ 1" X JS < M V JJ EH EH EH 0 CO W 66

More recently the differential coursework hypothesis was chal- lenged in a controversial research project on gifted math students by Benbow and Stanley (l980). These researchers indicated that the gender differences in SAT-Math scores are not due to differences in math preparation alone, but to differences in the mathematical reasoning abilities of boys and girls. Their results were challenged by Luchins and Luchins (l98l), Stage and Karplus (l98l), and Tomizuka and Tobias (l98l) and counter-challenged by Benbow and Stanley (l98l). A more pragmatic research resolution to the problem was presented by Pallas and Alexander (l983) based on analysis of a representative group of 6,000 students who took both the SAT-Math as twelfth-graders and the SCAT-Math in the ninth grade. Their study indicates that females score somewhat lower than males on the SAT-Math, largely because they are less likely to enroll in quantitatively-oriented courses (geometry, trigonometry, calculus, and physics) . However, highe-r grades in math courses of women than of men were a positive factor. Other factors—socioeconomic status, race, and gender—had small but statistically significant influences on SAT-Math scores. They conclude: We find that the male-female gap in SAT-M performance shrinks considerably when sex differences in quantitative high school course-work are controlled. These findings suggest that increasing females' rates of enrollment in high level mathematics courses would greatly reduce the sex difference in quantitative SAT performance and that it is premature to reject socialization and experimental explanations for the male-female gap in levels of quantitative performance. . . . Although the present research, and that of others, indi- cates that sex differences in mathematics performance arise during the high school years, we believe it is quite likely that many of the mechanisms responsible for the male-female gap in performance are set into motion much earlier, perhaps even in the elementary grades. (Pallas and Alexander, l983: l65, l80) The general consensus is that these gender differences in college admission math test scores can be largely accounted for by the differ- ences in the amount of mathematics, physical science, and computer programming courses that high school and college-bound women take com- pared to their male peers. Also, some evidence suggests that prior to and during college, women—including those who major in engineering and science programs—are less likely to engage in extracurricular and personal activities that might enhance either quantitative or physical science achievement (e.g., computer camps and science fairs) or mechanical and problem-oriented activities (e.g., reading Popular Mechanics, doing construction, and acquiring electronic hobbies). 67

TABLE ll: College-Bound Students' High School Preparation and Grades Preparation in Years l97l None l 2 3 4 5+ Mean Mathematics Men 0 2 ll 29 50 9 3.50 Women 0 4 l4 40 33 4 3.l0 Physical sciences Men l3 3l 37 l7 2 0 l.60 Women 24 45 25 5 0 0 l.l0 l975 Mathematics Men 0 2 l0 27 5l l0 3.57 Women 0 3 20 37 35 5 3.l7 Physical sciences Men 9 34 34 l7 5 l l.80 Women l5 43 29 9 4 0 l.45 l98l Mathematics Men * 2 8 23 54 l3 3.68 Women * 2 l5 32 43 8 3.38 Physical sciences Men 6 26 36 24 5 2 2.0l Women ll 38 34 l4 2 l l.59 l984 Mathematics Men * l 7 20 56 l6 3.78 Women * 2 ll 28 49 l0 3.34 Physical sciences Men 6 25 37 25 6 2 2.05 Women 9 37 36 l6 3 l l.69 l985 Mathematics Men * l 6 20 58 l5 3.80 Women * l l0 28 5l l0 3.58 Physical sciences Men 5 24 38 27 6 2 2.08 Women 8 34 38 l7 3 l l.74 *Less than l percent, but not zero. SOURCE: Educational Testing Service National College-Bound Seniors, i l985. _l 68

in Mathematics and Physical Science, by Sex, l97l-l985 (in percent) Grade in Course l97l A B C D F Pass Mean Mathematics Men 26 35 28 9 2 0 2.70 Women 26 36 28 8 2 0 2.80 Physical Sciences Men 27 39 27 5 0 2 2.40 Women 26 39 26 5 0 3 2.40 l975 Mathematics Men 28 38 26 6 l 0 2.87 Women 27 34 27 7 l 0 2.84 Physical Sciences Men 33 43 2l 3 0 0 3.05 Women 30 45 21 3 0 l 3.02 l98l Mathematics Men 27 39 27 6 l * 2.85 Women 26 39 28 6 l * 2.83 Physical Sciences Men 29 42 25 4 l * 2.94 Women 28 48 25 4 l * 2.94 l984 Mathematics Men 27 39 27 6 * * 2.86 Women 26 39 28 6 l * 2.84 Physical Sciences Men 28 4l 25 4 l * 2.92 Women 27 43 25 4 l * 2.92 l985 Mathematics Men 27 39 27 6 l * 2.85 Women 26 39 27 6 l * 2.87 Physical Sciences Men 28 4l 26 4 l * 2.9l Women 27 43 26 4 l * 2.9l Princeton, N.J.: Admissions Testing Program, The College Board, l97l- 69

TABLE l2: SAT Score Averages for College-Bound Seniors, by Sex , l970-l985 Verbal Mathematics Year Male Female Male Female l970 459 46l 509 465 l97l 454 457 507 466 l972 454 452 505 46l l973 446 443 502 460 l974 447 442 50l 459 l975 437 43l 495 449 l976 433 430 497 446 l977 43l 427 497 445 l978 433 425 494 444 l979 43l 423 493 443 l980 428 420 49l 443 l98l 430 4l8 492 443 l982 43l 42l 493 443 l983 430 420 493 445 l984 433 420 495 449 l985 437 425 499 452 SOURCE: Educational Testing Service, National College-Bound Seniors, Princeton, N.J.: Admissions Testing Program, The College Board, l970-l985. Women and Computers An analysis of factors having an impact on enrollment of women in engineering and science fields cannot ignore the increasing differences in the precollege computer backgrounds of male and female students. Armstrong (l986) reviewed the current literature and found gender dif- ferences in computer access, usage, and interest: in order to develop a background in computer programming, students need access to computers. Gender differences in computer access and usage have been of critical interest during the past five years. A California state-wide assess- ment of computer literacy at the sixth and twelfth grades concluded that girls either have or take fewer opportunities to work on computers at both home and school than do boys. More often than girls, twelfth-grade boys are inclined to learn about computers at home from friends or through the use of video games. More sixth-grade boys than girls are likely to have home computers. Even though both male and female students reported learning about computers at school, females exhibited a pattern of less participation and access than 70

that of males (Fetler, l985). Boss (l982) reported that girls did not have the same amount of time on computers in schools because they did not take the time to use the machines. These findings concur with Sheingold, et al., (l983), whose research indicated that gender differences in computer use spanned all grade levels, with males having higher amounts of usage and access. At the secondary level, computers were polarized in the mathematics/science and business areas, where women students tended to use only the machines in the business area (Sheingold, et al., l983). Several studies investigated gender differences in com- puter camp registration. Three times more males than females registered in computer camps with differences more pronounced at the secondary level than at the elementary level (Hess and Miura, l985). Gender differences in computer access and usage were found in the homes of students as well. Fathers tended to bring the computer into the home and teach others how to use the machine. In addition, families were more likely to buy a computer and spend more money on computer software and peripherals for male children than for female children (Miura and Hess, l984). At the precollege level, more male students than female students indicated interest in computer programming courses (Becker, l982; Hess and Miura, l985; Fetler, l985). The gen- der differences widen at the more advanced levels of education (Lepper, l985). In computer camps, merely five percent of the girls took the assembler language courses (Hess and Miura, l985). Lepper reported similar discrepancies in participation in the advanced courses in programming. Linn (l985) reported that women constituted 86 percent of the students in word processing, but only 37 percent of those in computer program- ming. Results of a Computer Literacy Scale administered to beginning engineering freshmen at Purdue University in l985 and replicated in l986 indicated that significantly more males than females had written and debugged programs, listed a program, saved a program, renamed a program, used a program to solve a mathematics problem, modified someone else's program, instructed others in the use of a computer, documented a program, used a word processor, and used a computer at home. The only item on the scale that did not indicate significant gender differences was "used a computer at school" (Armstrong, l986; Armstrong, et al., l985; Armstrong, et al., l986). Furthermore, the male engineering students entering college had completed more weeks of high school instruction in computer programming and knew more computer languages than their female counterparts (Armstrong, et al., l986). Several researchers have attributed the gender differences in com- puter background to differences in interests and attitudes. Armstrong (l986:26-27), in her survey of the literature, reported the following: 7l

A number of studies have found that men indicate more interest in computers than do women (Lockheed, Nielsen, and Stone, l983; Hess and Miura, l985; Campbell and McCabe, l984; Fetler, l985). Differences in interest in computer programming have also been documented (Becker, l982). However, once men and women were enrolled in the same computer programming course, their levels of interest tended to be similar (Schubert, l984). Research has not been conducted that is designed to identify whether interest extends beyond the duration of the course. In addition, differences in interest in computers do not extend equally over all grade levels. More significant gender differences in interests in computers at the sixth- grade level were found than at the twelfth-grade level (Fetler, l985). Self-efficacy (self-appraisal of one's ability to suc- cessfully complete a task) has been proposed by Bandura (l982) as a factor that may determine whether a person will engage in a task. In the computer field, Miura and Miura (l984) re- ported that male students had significantly higher perceived self-efficacy for computer-related tasks. These differences in self-efficacy were also reported among junior high school adolescents (Miura, l984). Women engineering students, in general, reported lower self-perceived abilities than did male students (Linden, et al., l985). Self-perceived abili- ties could help to explain gender differences in enrollment in computer programming courses. Researchers have proposed a myriad of reasons for the observed gender differences concerning interests and attitudes towards computers. The reasons include sex stereotyping, which begins in the second grade (Stein and Smithells, l969), traditional differences in mathematics and science (Winkle and Mathews, l982; Lockhard, l980), the placement of the com- puter in the mathematics and science departments (Saunders, l979), sex differences in spatial abilities (Aiken, l972), public stereotyping of the computer as a man's machine (Hawkins, l985), and the male-orientation of the major pro- portion of available software (Miura and Hess, l984). All of these reasons, and many more, can help to explain the docu- mented differences in interests and attitudes. If male students tend to possess higher levels of computer back- ground and interest, do they out-perform female students in computer programming courses? In general, research findings indicate that there is no difference between the academic performance of male and female students in computer programming courses, even though computer back- ground differences may be present at the beginning of the course (Armstrong, l986; Armstrong, et al., l986; Campbell and McCabe, l984; Mandinach and Fisher, l985). One contrary finding in a study concerned with teaching LOGO programming to elementary school children was that boys exhibited higher levels of academic achievement than did girls (Pea, et al., l984). 72

To summarize, there seems to be evidence, at least in the precol- lege years, that males have higher levels of interest in, access to, and participation in computer-related activities including games, com- puter camps,'and computer programming courses than females. In gen- eral, there are no gender differences in academic performance in com- puter programming courses at the precollege or college level. Linn and Dalbey (l985:202) summarized their "quasi-experimental" study of various computer-intensive sites by stating that gender differences, not evident in performance, were shown in interest in enrolling in computer courses: "The point where equity seems most central is in securing the participation of females in programming courses." Comprehensive Studies Related to Undergraduate Participation of Women in Science and Engineering A number of studies that provide comprehensive views of the factors influencing the underrepresentation of women in science and engineering have special relevance at the undergraduate level. Five major ones cited below are as follows: Engineering Education and Practice in the United States (National Research Council), An International Perspective Regarding Women in Engineering (Australian Bureau of Labor Market Re- search), Impact Analysis of Sponsored Programs To Increase the Parti- cipation of Women in Careers in Science and Technology (Denver Research Institute), Women in Engineering [National Center for Higher Education Management Systems (NCHEMS)], and Women in Engineering (Georgia Insti- tute of Technology). In addition, information is provided about still other studies of women's participation in science and engineering in other sections of this report. Engineering Education and Practice in the United States (National Research Council) In l980, NSF asked the National Research Council to conduct a study of the state and future of engineering education and practice in the United States. During the next five years, a committee of the Coun- cil's Commission on Engineering and Technical Systems conducted compre- hensive studies and analyses. Of special relevance to women in engi- neering and science were the final report and the panel reports on undergraduate engineering education, graduate education and research, and infrastructure diagramming and modeling (National Research Council, l985a, b, c). The latter report documents the tenfold increase in freshman enrollments and the seventeenfold increase in B.S. degrees awarded to women, indicating that the rapid increase in numbers of B.S. degrees in engineering came from two major sources, incoming freshmen and transfers from other fields and institutions. The undergraduate panel's central observation was that two factors central to the supply of women are institutional differences and field differences (National Research Council, l985a:24): In l982 the percentage of B.S. degrees awarded to women 73

from the 50 institutions having the largest number of under- graduate engineering students ranged from a high of 29.5 per- cent (General Motors Institute) to a low of 8.9 percent (Iowa State University). The largest number of B.S. degrees awarded by one school to women in l982 was 203 (l5.6 percent) from Texas ASM University, which graduated the largest total engi- neering class that year. The numbers also vary across engi- neering fields: in 1982, 29 percent of industrial engineering students (the highest percentage) and 24.5 percent of computer engineering students were women, while l0 percent of mechani- cal engineering students (the lowest percentage) and l3 per- cent of civil engineering students were women. The report also calls attention to two major curricular factors related to the preparation of women for engineering: The percentage of women in engineering programs appears to have no inherent limit. There are as many young women as men in high school who study mathematics and science through trigonometry and chemistry. However, almost twice as many young men as women take high school physics, calculus, and introductory computing. Other observations of relevance in the undergraduate panel report include the following: Apparently an interest in physics is an important factor leading to a career in engineering; men are attracted to engineering mainly by taking high school physics, while women are attracted to engineering through chemistry and biology. High school women often feel tracked away from physics; very few physics teachers are women, and course content and qual- ity are quite variable, often not appealing to women. Edu- cational experiments indicate that nontraditional approaches to the teaching of both physics and introductory computer subjects in sex-balanced classes result in their increased appeal to women students. The increased number of women students has helped make engineering schools a more attractive environment for them. Despite recent improvements, however, women students will report feelings of isolation, lack of acceptance by faculty and male student peers, and lack of acceptance of their ca- reer goals by friends, family, and their universities. Many women students still find engineering schools to be stressful environments, and they need support to help them deal with the difficulties that they encounter. But they do not form a homogeneous group, and their needs vary. For example, some report significant problems in adjusting to a strongly male environment; some find a supportive environment in a particular department; and many find support in a confidant, sometimes a close male friend. Some of these are problems 74

that will lessen over time as the number of both women stu- dents and women faculty increases. While increased use of foreign nationals as graduate teaching assistants and as faculty members is often cited as a problem because of language barriers, the practice also brings special problems for women students. Anecdotal evi- dence suggests that students and faculty from cultures in which the role of women is subservient may not be sensitive or sympathetic to the career aspirations of American women engineering students. . . . The Panel on Undergraduate Engineering Education recom- mends that, to achieve the full potential that this human resource offers, colleges of engineering, school systems, government, industry, and the engineering profession continue to work to increase the number of qualified women who study for a career in engineering. A key requirement is the need to encourage the study of mathematics and science by female secondary school students. (pp. 24, 26) This final report in the Research Council's comprehensive engi- neering study includes especially relevant observations (l985a:6-8): Since the early.l970s, considerable effort has been de- voted to increasing the participation of women and minori- ties in engineering. The recruitment efforts have paid off: the percentage of minorities in the engineering work force has doubled and the percentage of women has more than tripled. Currently, more than l5 percent of engineering undergraduate students are women (as compared to about l percent in l970), which has generated a feeling of success among many of those concerned with the issue. However, some sobering facts should be pointed out. Compared with the sciences and other professional disci- plines, women are still a small part of the engineering work force. Perhaps even more significant, beginning in l982 there has been a mild slowdown in enrollments of women in engineering. . . . The committee believes that the determination of appro- priate levels of representation in engineering for both women and minorities is not a matter for judgment by panels of educators and industry representatives. These are social questions requiring broader discussion. However, both women and minorities are represented as students and as practi- tioners in engineering at lower levels than in other science and technology professions. Therefore, the committee con- cludes that the participation of women and minorities in engineering should be matters of continuing concern to the engineering community. There is still much to be done. A case in point is the treatment of women on engineering faculties. There is a recurring perception of bias against female faculty members in assignment of teaching responsi- 75

bilities, in selection for research teams, and in granting tenure. In many schools there also appears to some to be a bias against female graduate students as candidates for faculty positions and in the provision of financial and intellectual support. College administrators should make a candid assessment of the attractiveness of academic life for women in their institutions, and if negative aspects such as these are found, they should take firm steps to eliminate them. An International Perspective Regarding Women in Engineering (Australian Bureau of Labor Market Research) In spite of limited time and resources, Byrne (l985), in a study for the Australian Bureau of Labor Market Research, was able to col- lect, analyze, and synthesize considerable data on women in engineering from a wide variety of international sources. A number of her con- clusions and recommendations have special meaning and significance internationally: The male-dominated leadership of the engineering profes- sion, unlike most people, is now coming to recognize that technology is androgynous. . . . The first matter to be stressed is that, while the re- search and studies reviewed in this report provide a firm base for strengthened hypotheses and for choosing some policy options, they do not yet provide decisive or conclusive evi- dence on the causes of female under-recruitment or success in engineering. Few have looked at longitudinal effects or trends. Secondly, most studies are either macro and statis- tical where not enough is known about the cohort's personal characteristics, or micro in detail and not always wholly transferable. Nor is enough known about actual sex differences in sci- ence and technology students. what is clear is that sex dif- ferences between successful male and female students—those who enroll—are often minor. But it is not clear what the key differences are for those who do not enroll. . . . Many questions are still unanswered and require further research. Research has not yet produced adequate answers about why women choose some branches more readily than others. . . . The importance issue for further enquiry are those of institutional and discipline variations, two factors which are common in all countries reviewed (whatever their educa- tional structures). Further research is needed to explore these. Respondents were asked about the profile of women engi- neers. Some of the data collected are encouraging—their equal ability, high achievement, short or non-existent child- rearing gap and so on. Again, however, the data do not exist 76

in readily available form to support firm predictors yet of "successful" and "unsuccessful" profiles. In summary, Byrne noted the significant growth in the number of women engineers since l974. Positively influencing the growth were "institutional commitment, increased 'carrots' or 'sticks' by way of grants or special interest group pressure, manpower and womanpower shortages, and new fields opening up (computing and biomedical) which are less male-oriented from the start." in addition, she cited the following negative influences: • Refusal to accept responsibility for the problem, or acceptance only if (a) women do not compete with men for rationed places and become a threat and (b) institutions do not have to real- locate existing resources; • Refusal to accept that "the best man for the job may be a woman"; • Lack of leadership from the professional institutions; • Adverse public image of engineering; • Lack of resources from institutions, government grants, or industry to fund new programs; • Lack of role models or use of them; • Need to educate parents and teachers and bring them into the twentieth century in terms of changing sex roles; • Male-oriented, biased, out-of-date or non-existent career ad- vice, usually also too late to influence or save curricular choices; • Career influence in senior high school is too late (interven- tion should start in junior high school); and • Deficiencies in math and physics. An Impact Analysis of Sponsored Programs To Increase the Participation of Women in Careers in Science and Technology (Denver Research Institute) Another related study was conducted by the Denver Research Insti- tute in order to evaluate and to assess six experimental programs re- lated to those projects: those at the University of Kansas, Queens- borough College, the University of Missouri at Kansas City (UMKC), Rosemont College, and Massachusetts Institute of Technology and a pro- gram originally developed under an Institute for Educational Develop- ment grant and completed by the Policy Studies in Education group in New York. A number of conclusions and recommendations emerged from that re- port including the following (Lantz, et al., l976:l-2): The first recommendation is that given the emphasis on professional careers in science, concentration on high ability women is a realistic restriction. Second, the more able and/or more mature women appear to be more interested in con- tent of the vocation, while younger or less able women are 77

more interested in lifestyle possibilities relevant to any career. A trade-off between age and ability was noted; and increase in ability level may decrease the appropriate age. The third recommendation was that the programs concentrate on women who have already shown an interest in science and have the prerequisite background after the junior year in high school. It was also recommended that a distinction between career education and programs to interest women in science- related careers be made. The general conclusion was that re- cruitment and commitment should be emphasized in high school years; reinforcement, support and retention be emphasized in the college years; and that removing institutional barriers should be emphasized in graduate school, reentry programs for mature women, and post-employment programs. It was concluded that most of the participants felt strongly that all-female workshops and classes were necessary under certain conditions. Another general conclusion was that role models appeared to be the most effective component of most of the projects. It was recommended that a mixture of role models, closer in age and accomplishment level to the participants, be utilized in conjunction with inspirational models, successful women at the top of their fields. It was also suggested that reentry programs might be more successful in recruiting underemployed, rather than unemployed, women. Another recommendation was that the National Science Foundation develop a strategy to disseminate the curriculum materials of each of the projects. It was also suggested that the Foundation take the responsi- bility for making evaluation instruments of established reli- ability and validity available, to the project directors and for providing technical assistance in their internal evalua- tion efforts. A conference of all federal agencies involved in career education was also recommended to provide a means to pool available resources, and to avoid duplication of ef- fort. Finally, it was recommended that the Foundation con- tinue its activities in encouraging women to choose science as a career and to disseminate knowledge about the existence of its programs. The fourth chapter in the report discussed alternative interven- tions. It began with a list of psychological, sociological, and institutional barriers to the participation of women in science-related careers. Some assumptions were delineated, and different kinds of interventions or treatments were proposed. Each of the possible interventions was categorized by educational level (e.g., elementary school, high school, college, graduate school, reentry, and postemployment programs). Lantz and her colleagues stated four major assumptions and provided specific interventions for each. As possible interventions to the assumption that "knowledge that science-related careers are open to women is a prerequisite for pursuing those careers," the following were suggested: 78

• Workshops and seminars portraying professional women in science careers. Several NSF-sponsored projects are imple- menting this approach; for example, Mary Baldwin College was funded for seminars in l975, and many more have been funded in l976. These workshops might encompass three components of discussion by or with the role models: actual job con- tent, lifestyles, and on-the-job problems. Depending upon the interests of the participants, the various aspects could be differentially emphasized. For many science majors, the different job titles that may be pursued from a science major may be of the most interest. For other groups, life- style preparation and solutions of on-the-job problems (if any) may be of more interest. • Increase in the number of female science professors. Even though the majority of women having careers in science .spe- cialize in teaching, the percentage of full' professors is very low in the sciences. The most obvious and most avail- able role model for women interested in science would be their professors. In addition, it is likely that female professors may be more supportive of female undergraduate majors than male professors. Therefore, programs to in- crease the number and status of female professors are en- couraged. The programs to increase the number may encompass all of the programs suggested in this section, and there may be many years before results are observed. Direct support of programs to increase the status of current female pro- fessors may produce results in a shorter time period. Many such programs are referenced in the section on_programs for women currently in the labor force. (p. 69) A second assumption was that "career-committed females may benefit from special counseling, support group activities, and other forms of social encouragement" (Lantz, et al., p. 69). Two possible interventions were given: • Special counseling or support groups for women planning to pursue nontraditional science careers. These special coun- seling or support groups may take many forms. The groups might be task-oriented around special courses that would interest primarily females or be more social in nature. Whatever the vehicle used, the major aim would be to de- crease the alienation and social pressure by encouraging friendships among women with similar values and aspirations. • Special housing for women planning to pursue nontraditional science careers. While special housing presents many dif- ficulties at large universities, group housing has been shown to increase the retention rate of female engineering majors. Wherever this may be an option, different housing arrangements may be tried. (pp. 69-70) In addition, Lantz, et al. (l976:70) noted that "many fail to pur- 79

sue science-related careers because they fail to successfully complete prerequisite mathematics courses." They suggested that remedial mathe- matics courses, innovative teaching methods, and special tutorial pro- grams might encourage interest in scientific careers: Offering "remedial' mathematics courses or special courses, such as the UMKC project, may be an appropriate vehicle to assist women who are interested in science but have diffi- culty in advanced courses because of inadequate backgrounds in math. Even for those women successfully completing mathematics at the high school level, innovative approaches to teaching mathematics could be applied to advanced courses, such as solid geometry and calculus. For women taking ad- vanced mathematics courses in college, special tutorial pro- grams, run by other women, might assist their colleagues. The women interviewed at the University of Kansas expressed their shyness with male tutors and reluctance to ask the teachers for help. They felt that not understanding a single lesson usually meant it was impossible to comprehend any subsequent lesson. Therefore, they wanted female tutors who were immediately available and consistent. Finally, this study noted that because "tangible and intangible institutional barriers discourage women from pursuing science-related careers" (p. 70), colleges and universities should take specific steps to alter the situation: • Increasing the number and percentage of women holding undergraduate assistantships in teaching and research in the sciences. A "spin-off" of at least one of the experimental projects (not currently complete) seems to be the very positive effects of hiring junior and senior science majors as staff personnel. It not only served as a financial aid, but also it was interpreted as a "vote of confidence" and served to increase interest, exposure, and expertise in their areas. Assistantships also provide additional encouragement to go to graduate school and usually provide a closer relationship with a faculty member. Assistantships to declared science majors may improve the retention rate and result in more women attending graduate school in science. • Increasing the number and percentage of women in science- related cooperative and intern programs, as is now at- tempted by one of the current NSF-sponsored projects. Such programs should result in a better understanding of job options, job requirements and preferable job alternatives (discovering that one doesn't like a job is as important as discovering that one does). Further, it may provide better "connections" to obtain a job or to gain admission to graduate school. 80

• Rewriting graduate "fellowship" brochures. One discouraging factor in applying for graduate school is the way informa- tion on financial aid is presented. The brochures, espe- cially on the most prestigious "fellowships," are uniformly written in masculine gender and appear to rule out women. • Increasing the number and percentage of female science pro- fessors (advisers). (pp. 70-7l) Women in Engineering: An Exploratory Study of Enrollment Factors in the Seventies (National Center for Higher Education Management Systems) A special study under a National Institute of Education contract (Corbett, et al., l980) examined general trends in enrollment, attri- tion, and degrees granted as well as case studies of six institutions (Vanderbilt, University of Washington, Purdue, Colorado School of Nines, Prairie View, and New Mexico State). It concluded that external causes and conditions were the primary factors that influenced the in- creased participation of women in engineering education and that in- ternal influences were primarily facilitative. External factors that seemed to lead to the increased enrollment of women in engineering programs during the past decade include the following: • Job market outlook for engineers generally and opportunities and competitive salaries for women in particular, • Industry recruitment of women for engineering positions and industry support of engineering-education programs through funding of scholarships, co-op programs, summer employment, and campus career institutes for prospective students, • A change in the image of engineering as a field now central to public-policy and environmental issues, • A change in society's attitude toward women's roles in the work force and the consequent change of women's attitudes and awareness regarding engineering as a career, • Federal financial-aid programs for students, • Affirmative-action pressure on industry, • High school teachers and counselors encouraging students to acquire academic backgrounds and interests relevant to engi- neering studies, and • Parental and family influence on the role perceptions of women students and their choices of college or university. The NCHEMS study stressed that these external forces do not operate in isolation, but instead interact synergetically to create both con- straints and opportunities. Similarly, engineering schools and their universities "are lodged in a complex network of formal organizations that influence each other" (p. l2l), including both the businesses and industries hiring their graduates and high schools from which the engineering students traditionally come. 8l

Internal factors were also cited by Corbett, et al. (l980:l2l-l23), as influencing increased women's enrollment in engineering programs: • Enrollment declines, which create a new awareness of women as a potential new pool of students, • Recruitment efforts to make high school women aware of opportunities for them in engineering, • Persistent efforts to increase the number of women in engineering programs, • Support from both administrators and faculty, • Designating responsibility for coordinating efforts to attract and support women in engineering programs, • Student organizations that provide social support and professional affiliations for women students in engi- neering, • Scholarships for incoming and returning women students in engineering, • Retention activities directed to the particular needs of women students, and • Location of the institution. While many of these efforts were not specifically intended to help women, prevailing societal forces made larger numbers of women more likely to take advantage of them because (l) they were high achievers in high school, and (2) they came from families and socioeconomic backgrounds such that a choice of an engineering major was a relatively minor shift from a choice of science or math in college, which they might otherwise have made. Corbett and his colleagues noted, "The particular coincidence of the efforts of engineering schools to recruit more students and the readiness of more women to entertain nontradi- tional careers probably explains why many schools felt that the in- crease in the numbers of women appeared to "just happen,1 without extraordinary effort on their part" (p. l23). Other, more subtle, influences were also identified in the NCHEMS study as factors influencing the increasing representation of women in science and engineering: • Available information about students (grade-point average), prospective students (SAT and ACT scores), and programs (retention rates), • Federal initiatives, such as direct financial aid to stu- dents, funding for research and demonstration projects that have given an impetus to women's enrollments in engineering (for example, the NSF and HEW grants to New Mexico State University and the Women's Education Equity Program Grant to Purdue University), legislation that legitimizes support foe affirmative action and puts pressure on industry to re- spond, contracts with industry to develop or produce goods and services that keep the job market for engineers favor- able, and compilation of data that indicate the relative positions of women in engineering and the sciences, and 82

• Activities of professional associations, such as support to student engineering organizations, national meetings and workshops, publications that express concerns about women in engineering, and information exchange networks and clear- inghouses (such as the Women's Educational Equity Act Pro- gram Network) for research on and exemplary practices con- cerning women in the professions. The authors also noted that increasing the enrollments of women in engineering programs has required no changes in curriculum, basic admission policies, or quality standards. They concluded, Indeed, the expansion of the applicant pool by attracting more competitive women has apparently helped maintain high overall admissions standards during periods of decreased en- rollments. The main enterprise of engineering education has not been changed, either to stimulate or to accommodate women students. Faculty resistance has not been an issue, pre- sumably because little or no change was required of most fac- ulty. The large majority of the women entering engineering programs have been good students, and programs for recruiting and supporting them rarely threatened existing resources. The composite picture formed by these case studies shows the institution reacting without resistance to a confluence of environmental forces conducive to the increased enrollment of women in nontraditional programs. . . . (pp. l25, l26) In other words, as the numbers of women increase, some of the barriers to women implicit in an environment where they are a small minority are lowered without conscious institutional effort (Kanter, l977). Women in Engineering: Policy Recommendations for Recruitment and Retention in Undergraduate Programs (Georgia Institute ofTechnology) This study, conducted under a grant from the Fund for the Improve- ment of Postsecondary Education, focused primarily on recruitment, although some attention was also given to retention. Thirty of the most successful and 30 of the least successful programs in attracting women into engineering were examined, and an in-depth study of one in- stitution (Georgia Institute of Technology) was conducted. Connelly and Porter (l978:4) concluded that two factors accounted for a substan- tial part of the institutional differences observed: 'larger schools have tended to do better than small ones, and academically or socially elite schools have outperformed less-elite schools." They identified what they considered to be the two key underlying mechanisms: • The Decision Support Hypothesis: Women considering engi- neering as an undergraduate major are doing something unusual. They are unsure about what the study or practice of the profession will involve: family, friends, and 83

teachers will generally have little helpful advice; there are few women engineers with whom to talk; and the situa- tion is changing rapidly. In short, engineering is a "high uncertainty" choice for a woman, and relatively more reassurance or "decision support" is needed to reduce this uncertainty. The importance of support programs for women already in engineering school is documented by Davis (l965) and others. We are here arguing for a more general hypothe- sis, that support is needed before entry as well as dur- ing undergraduate training. The nature of this support will vary over time and by situation. Prospective stu- dents can benefit from perception that they are welcome and will get a fair chance. New frosh need campus-oriented support; seniors, job-oriented reassurance. • The Positive Feedback Hypothesis: It appears that one of the most helpful factors in attracting women is to have a sizable number already. This helps in various ways: social and academic support groups on campus, involving both peers and more advanced students (often referred to as having a "critical mass" of women students); establishing a grapevine for course and career guidance; "role models," both older students and women faculty; informal recruiting, through contact with current high school students; attitude change, as high school and college faculty and counselors see women succeeding as engineering students; and so on. (pp. 5, 6) Enriched Environments and Background Study An interesting sociological approach to examining the enrichment factors associated with selection of nontraditional fields by women was conducted by Carol Auster (l984). She focused her Princeton doctoral dissertation under a Rockefeller Foundation/University of Illinois grant on men and women who were attending accredited U.S. engineering institutions. Her primary thesis was that "women in engineering would emerge from more enriched backgrounds than men in engineering," and her primary statistical tests were contingency table analyses and discrimi- nant analyses. The following exerpts from her thesis are especially relevant: To some extent, the hypothesis that the women would emerge from a more "enriched" environment than the men is supported. For example, women were more likely than men to have highly educated parents with high incomes, fathers in engineering, mothers with full-time salaried jobs, and have spent most of their life in an urban area. Women also re- ceived overwhelming encouragement from those deemed most important and parents' encouragement was affected by some of the aforementioned social background factors. Other aspects of women's initial decision and continued commitment to engi- 84

neering seemed to further reflect the pioneering or nontradi- tional nature of their choice, including perceptions of the attitudes and abilities of women and men in the engineering program, and interest in engineering specialties and potential employers. In the initial decision making phase, the support of family, particularly parents, and the immediate milieu and academic success in high school were very important; however, the new immediate milieu, peers and professors, and engineer- ing grade point average became important in continued commit- ment to engineering. (pp. xix-xx) . . . Despite the many unanswered questions about occupa- tional choice, many suggestions for those interested in re- cruiting more women into engineering are implicit in the find- ings of this research. Women must continue to take math and science courses throughout high school and college, be- en- couraged in those endeavors, and hopefully emerge with the confidence to pursue further work in these fields. Since those women are most likely to be the products of "enriched" backgrounds, educators could carefully tap into that group or further encourage the less likely candidates as a way of ex- panding the pool of potential women engineers. Further in- terest in engineering seems to be sparked by contact with those in the field. This suggests that science and engineer- ing conferences which put women into personal contact with employed scientists and engineers should be continued, par- ticularly if close relationships could be established between the working engineers and those potential engineers. Once a woman decides to enter an undergraduate engineer- ing program, she should be surrounded by people who offer support for such a decision and by professors whom she per- ceives rate the sexes equally for potential success in engi- neering. Furthermore, interaction with both male and female peers, especially discussions about engineering careers, seems to have a positive impact on these women. The field of engineering may be shaped by the new inter- ests brought to the field by the increasing numbers of women. Women seem particularly interested in people-oriented work functions and environmental issues. The interest in nature and the environment might help technologists recognize the need to work with nature, not conquer nature to the extent that the earth becomes uninhabitable. Eventually, men's and women's engineering specialties may reflect only individual and not learned gender-linked interests. (pp. 330-33l) Critical Mass, Role Models, and Unisex Programs A number of studies and analyses indicate that the socialization process is greatly facilitated if there are a sufficiently large enough proportion and number of students having similar, albeit atypical, characteristics (for example, women, blacks, foreign nationals, or dis- 85

abled). Somewhat related socialization factors are the importance of role models and the impact of special programs for target groups (pro- grams for women only or unisex approaches). Unfortunately, there have been very few experimental or quasi-experimental studies in these areas of the research, and evidence is anecdotal, correlational, and "de facto." Critical Mass. The "critical mass" theory has been proposed by a num- ber of investigators (Byrne, l985; Connelly and Porter, l982; Lantz, et al., l976). Although opinions vary considerably on what proportion or number constitutes a critical mass, the primary thrust of the argu- ment is that until a sufficiently large enough number or proportion of women are enrolled in an institution, program, or course, the social- ization process constitutes and remains a barrier. The resulting sex- role conflicts result in problems and hardships for women in science and engineering programs, which are inherently already difficult and demanding. However, when a critical mass is reached, many of the socialization problems are minimized, women recognize that there are others "in the same boat," and the resulting self-support groups are eventually institutionalized. Role Models. The "role model" factor, a very common element in most programs designed to attract and retain women and minorities in science and engineering, is based on the premise that women can learn from other women "who have made it" or "are making it." Although early role model efforts focused on "stars" (women with unusually high achievement or endurance records), more recent role-model efforts have broadened the base considerably to include more "average" and even "struggling" women who have or are pursuing education and careers in science and engineering. A general consensus among those that have developed spe- cial programs is that role models are of critical importance. However, their impact on women's decisions to obtain science and mathematics educations was challenged recently in a report from the Federation of Behavioral, Psychological and Cognitive Sciences, which is appended to the report of the National Science Board's Commission on Precollege Education (l984:l09): Research also has identified factors that influence stu- dents in selecting courses in mathematics and science, partic- ularly among young women and members of minority groups. In addition to achievement in previous educational experiences, individuals are more likely to continue their mathematics and science education if they perceive these fields to be relevant to careers that are available to them, and if they have gen- eral interest in "things," rather than primarily in "people." Contrary to some popular belief, exposure to female role mod- els in science and mathematics has been found to have little effect on young women's decisions to obtain science and mathe- matics education. Perhaps some middle ground is in order i.e., programs should make use 86

of a wide range of role models (not only women "stars," but average women achievers, dual-career couples, and male role models as well). Unisex Programs. The "unisex or majority role" approach reflects ef- forts to facilitate the socialization process by focusing on institu- tions, programs, and courses where women constitute the primary total participants. A substantial number of studies indicate that women's colleges have produced a larger proportion of women scientists than coeducational institutions. Many programs targeted to women in engi- neering, math, and the physical sciences were developed to make women feel more "comfortable" in "hands on," remedial math, or experimental science courses. Some experimental studies conducted on male-female ratios have pro- vided relatively unclear or nonconfirmatory evidence that same-sex or balanced-sex courses are most effective. One such study conducted at Purdue University utilized a "hands on" laboratory setting (one "all women" section and one "balanced" with equal numbers of men and women). No statistical differences in cognitive and affective variables were observed from both summative and formative evaluations. Anecdotal re- sults indicated that in the all-women's section, there was a "more re- laxed and less threatening atmosphere," but that there was a more "realistic" and "gender helping" atmosphere in the balanced-sex sec- tion. At the end of that study, the overwhelming majority of women and men, including participants in experimental and control groups, strongly endorsed integrated or "gender-mixed" lectures, laboratories, and counseling (LeBold, et al., l983). Retention of Women in Science and Engineering Programs Research on college attrition has generally indicated that cogni- tive factors, especially early college grades, are the primary factors affecting retention rates. In addition, a significantly large number of studies indicated some gender differences. Generally, these have indicated that academic achievement may be somewhat more relevant to retention and attrition for men and personal factors more relevant for women. This section focuses on two comprehensive national studies of engineering retention and one comprehensive institutional data base. Female Engineering Students: Attitudes, Characteristics, Responses to Engineering This study, conducted under a grant from NSF, employed a longitudi- nal data collection rationale; however, it was funded only for two years and long-range institutional data were limited. Of special rele- vance was the data collected on engineering retention. Ott (l978) analyzed the retention rates • of women and men in engineering at 42 schools. She estimated that the retention rates in engineering until the beginning of the sophomore year were 73 percent for men and 68 per- cent for women; women were more likely to be internal transfers, and men were more likely to leave for academic reasons or as voluntary 87

withdrawals. She also found that academic achievement in high school was positively related to engineering retention for both men and women. Also, women who remained in engineering, compared to women who left, were more apt to have fathers who were college graduates, to be "Cau- casian," to have done homework in high school, and to have parents who stressed the importance of attending colleae. National Engineering Career Development Study This study examined a number of factors associated not only with the educational and career development of beginning men and women engineering students, but also with the educational and career devel- opment of engineering men and women after graduation. Of special relevance to this report was the data collected on engineering reten- tion under grants from NSF and the EXXON Education Foundation. Shell, et al. (l985), studied engineering and university retention at 20 U.S. engineering institutions as part of this study. They found that 55 percent of the female and 74 percent of the male first-year students in l98l were still in engineering in their third and fourth years. They also found that women were more likely to be internal transfers than were men (l9 percent versus l5 percent, respectively). The primary -factors positively related to engineering retention for both men and women were academic performance during the first year in college; SAT-Math scores; and self-perceptions of math, science, and problem-solving ability. These same variables were found to be posi- tively related to retention of men and women in engineering both when each gender was analyzed separately and when men and women were com- bined in a single analysis. Other factors positively affecting engi- neering retention included the selectivity of the institution attended, high Nuclear and General Engineering Interest Inventory Scores on the Purdue Interest Questionnaire (PIQ), high school rank, and precollege grades in math and science. Some factors negatively related to engi- neering retention were (l) job values that stressed job comfort, in- come, job flexibility, and routine work and (2) high Management Inter- est inventory scores on the PIQ. Purdue University's Studies of Engineering and University Retention Purdue University has been monitoring university and engineering retention using longitudinal data bases for more than 25 years. During the last two decades, the university has focused considerable attention on l0-year longitudinal studies of engineering retention (do first-year engineering students persist and eventually graduate in engineering?) as well as university retention (do engineering students persist and eventually graduate in engineering and/or non-engineering fields at Purdue?). Researchers have found that short-term attrition studies that examine engineering or university retention rates through the sophomore year are often misleading; even studies that examine graduation rates after four years (or eight semesters) grossly underestimate the graduation rates. 88

Figure 2 provides a graphical perspective on the importance of longitudinal data in examining university and engineering retention and graduation rates. Purdue's women engineering students who began their studies in September l976 were tracked using computerized student rec- ords every semester for l0 years (note that most university and engi- neering attrition occurs during the first and second years). Most of the engineering students who were still attending Purdue after five or six semesters (junior year) eventually graduated with baccalaureate degrees. Figure 2 also indicates that estimates of graduation rates after four years (or eight semesters) grossly underestimate the longer- term graduation rates. Stopping out instead of dropping out also is fairly common for an increasing proportion of students, and many stu- dents participate in special work study or cooperative programs that result in undergraduate time periods greater than four years (or eight semesters). Retention data for men and women who began their engineering stud- ies at Purdue are very similar (see Figure 3). A slightly lower per- centage of women than men is retained in engineering, but a slightly higher percentage of women who entered in engineering is retained in the university and eventually graduate. Tests of the significance of the differences in the percentages of men versus women who are retained in or graduate from engineering or who are retained in and graduate from the university support the null hypothesis of "no statistically significant differences." This rich data base, which includes samples annually ranking from l,000 to l,500 students who have been followed longitudinally for l0 years, provides the basis for the following observations: • University and engineering retention rates improved signifi- cantly during the l970s. • Engineering retention improved significantly due to improved ad- missions and initial course placement, special programs designed to improve retention, and grade inflation. • Cognitive variables are normally better predictors of engineering and university retention than non-cognitive variables including gender, socioeconomic status, interest inventories, personality measures, and attitude measures. • Precollege variables are of less value in predicting engineering and university retention than college variables. Multiple re- gression, discriminant analysis, and other multivariate tech- niques usually indicate that precollege variables have negligible predictive value when college variables such as grades are in- cluded. • Essentially the same variables predict engineering and university retention both when analyses are made separately for men and for women and when both groups are combined. However, women who transfer to nonengineering fields or withdraw from college usu- ally have higher college grades than men who transfer or with- draw. • Engineering retention seems to be primarily related to academic performance at the college level (grades) and, to some extent, 89

100 Univ Retention -•- Univ Graduation -•• Eng Retention -»• Eng Graduation 8 10 12 Semester 14 16 18 20 Figure 2 University and engineering retention and graduation rates of women, Purdue University, 1976. •Q- Univ Women -•- Univ Men •o- Eng Women -o- Eng Men 75 77 79 Year Entered Figure 3 University and engineering retention rates (after six semes- ters), Purdue University, 1971-1983. 90

interest in the field and career content. Economic factors re- lated to the supply and demand of engineers and scientists in particular fields also influence transfer and retention rates. Graduate School and Related Financial Aid Although this paper has focused on undergraduate programs, deci- sions made at the undergraduate level are often critical in preparing and encouraging science and engineering students to pursue degrees be- yond the baccalaureate level. Until recently, women were not encour- aged to the same extent as men to pursue graduate studies in science and engineering and to seek and to attain federal support. Data on graduate enrollment and support (NSF, l983) indicate that women are less likely to receive federal and institutional support than men and more likely to be self-supporting. Although considerable ef- forts have been initiated in recent years to encourage more women to apply for federal and university fellowships and assistantships, Na- tional Research Council data indicate that fewer women apply for such assistantships and that the selection process may require further re- view to insure equitable treatment. The selection process for graduate admissions, financial support, and awards relies primarily on Graduate Record Examination scores, rather than on college and major grade-point index, and the recommendation of male faculty, who may not support equitable treatment for women in science and engineering. Potential discriminatory practices must remain a matter of concern and certainly merit careful analysis and study. Although institution-wide programs directed at providing more equi- table graduate opportunities are important and necessary, more effec- tive programs within institutions may require more grass-roots efforts at the college or departmental level. For example, the College of Engineering at the University of California, Berkeley, recently pre- pared an affirmative action report on graduate education (Pister and Humphreys, l986). The report recommended a comprehensive approach that begins with recruitment but also recognizes the need for retention, evaluation, encouragement, commitment, and even advocacy. It documents the actions and policies required to meet the challenge to develop, implement, and institutionalize an effective graduate program for women and minorities. Major Findings and Implications The major findings and implications of this report on undergraduate education for women in science and engineering can be classified into two major areas: (l) programs and policies designed to increase the number and proportion of women in science and engineering and (2) areas of research and study that are needed to provide more definitive in- formation useful for developing programs and policies that will lead to more equitable opportunities for women in engineering and science. 9l

Program Elements Some of the factors influencing the career choices and career de- velopment of women can only be affected by slowly evolving changes in society and culture. However, significant impact on other factors has been accomplished through direct interventions aimed at the individual as well as indirect strategies aimed at teachers, counselors, parents, and employers. Programs have been developed at colleges and universi- ties to influence positively the number of women selecting and success- fully completing science and engineering programs. Directories and books (Aldrich and Hall, l980; Bogart, l984; and Humphreys, l982) de- scribe in some detail these programs, many of which have carried out extensive evaluation of individual interventions. In addition, a com- prehensive evaluation of women in engineering programs throughout the country is presently under way by Susan Schwartz at the Stevens Insti- tute. A variety of elements constitute effective programs designed to in- crease the pool of women in nontraditional fields where they are under- represented. Fragmented programs that focus on single issues are often ineffective; and a recruitment program that either ignores retention or only focuses on merit awards or an occasional career day are not likely to be- very effective. A comprehensive institutional program de- signed to increase the pool of women in engineering or computer science should include at least four program elements: (l) recruitment, (2) re- tention, (3) employment, and (4) evaluation. At the undergraduate level, for example, a recruitment program might include the following: high school visitations; education and career information; teacher- counselor conferences; merit awards, scholarships, and financial aid; summer programs; academic year programs; personal follow-up programs; and parent-student information sessions. Similarly, an effective re- tention program might include orientation programs, special courses and seminars, peer tutoring, recognition awards and events, exit inter- views, educational and career counselling, cooperative and summer job programs, and student organizations, e.g., Society of Women Engineers. Since the overall plan is to eventually increase the pool of women in science and engineering careers, an effective program at the under- graduate level might also focus on the following: • Employment: invitations to prospective employers, summer intern- ships and cooperative programs, job search and career days, re- sume preparation, interviewing techniques, job listings, and college placement services. • Graduate School: undergraduate teaching or research assistant- ships, graduate student and faculty mentorships, invitations for department seminars, graduate school information sessions, campus visits by faculty representatives from leading advanced degree institutions, and dissemination of information on graduate school admissions and fellowships. Comprehensive programs also have a synergetic impact, both within pro- gram elements and between program elements. 92

The final element in developing a comprehensive program at the un- dergraduate level would include an evaluation program that might in- clude pre- and post-surveys for special programs, exit and graduate interviews or surveys, follow-up studies of graduates, systematic in- terviews and surveys of students, experimental and quasi-experimental studies, student data-base information systems, cognitive and affective testing, and faculty and counselor feedback programs. It is important to note that the evaluation element should be an integral part of the overall program; individuals and staff responsible for evaluation should be included in the planning and operation of an effective pro- gram. Similarly, evaluation data collected for its own sake without providing important and timely feedback is of limited value, even though they may contribute to researchers' fields of expertise and have some merit in their own right. Future Areas of Research on Gender Equity in Science and Engineering Four areas of future research are especially relevant to programs designed to facilitate gender equity in science and engineering: (l) data bases, (2) admissions, (3) retention, and (4) education and career development program evaluation. Summary The following generalizations can be inferred from the material presented: • Significant increases in the pool of women who are qualified and interested in education and careers in science and engineering are needed. • Increases in the pool of qualified women interested in science and engineering can be achieved by a combination of societal, institutional, disciplinary, and individual efforts. • To achieve more equitable representation of women in engineer- ing and science, special attention at the undergraduate level should focus on engineering and the physical sciences, the two major areas where women are significantly underrepresented. Within engineering, the disciplines in which women are most underrepresented at the undergraduate level include three of the five largest and well-established disciplines—mechanical, elec- trical, and civil engineering—as well as some of the smaller but significant disciplines including aerospace, nuclear, and petroleum engineering. In the physical sciences, physics and astronomy seem to be the primary fields quiring attention. • Significant progress has been made ir he past decade in en- couraging a larger percentage of coll* -bound women to enroll in math and physical science in high school, but the gap be- tween women and men in precollege preparation remains large. A new area of increasing importance is the potential development 93

of gender inequities in the precollege computer backgrounds of women entering science and engineering. • There is some evidence that a critical mass can have a syner- getic impact that facilitates not only recruitment, but also retention, career choice, and career development. Role models and programs designed to meet the special needs of women are also effective. • Additional factors that are important in increasing the percen- tage of women in engineering and science are institutional in nature. Private institutions, traditionally black colleges, and selective institutions seem to have been more effective than public, predominantly white, and less-selective urban institu- tions. • Institutions with strong comprehensive and administratively sup- ported programs that stress not only recruitment, but also reten- tion, graduate school, and science and engineering careers seem to be more effective than programs that focus on only one or two areas and are not supported by administration and faculty. The development of comprehensive institutionalized programs may be the key element necessary to ensure full participation of women in science and engineering. Bibliography Aiken, L. l972. Biodata correlates of attitudes toward mathematics in three age and two sex groups. School Science and Mathematics 72:386-395. Aldrich, M. L., and P. Q. Hall. l980. Programs in Science, Mathematics and Engineering for Women in the United States: l966-l978. Wash- ington, D.C.: American Association for the Advancement of Science. American Council on Education. l97l-l985. The American Freshman: Na- tional Norms. An annual report. Los Angeles: Office of Research and Cooperative Institutional Research Program, University of Cali- fornia. Armstrong, P. S. l986. Gender Differences in Predicting Academic Per- formance in Entry-Level College Computer Programming Courses. Master's thesis, Purdue University. Armstrong, P. S., W. K. LeBold, and K. W. Linden. l986. Predicting achievement in beginning-level computer-programming courses. In Proceedings of the l986 Frontiers in Education Conference. Arlington, Texas: American Society for Engineering Education. Armstrong, P. S., W. K. LeBold, and S. Ward. l985. Profile of the l985 Beginning Purdue Engineering Freshman. West Lafayette, Ind.: Pur- due University Department of Freshman Engineering. Astin, A. W. l977. Four Critical Years. San Francisco: Jossey-Bass. Auster, C. l984. Nontraditional Occupational Choice: A Comparative Study of Women and Men in Engineering. Princeton, N.J.: Princeton University Press. Bandura, A. l982. Self-efficacy mechanism in human agency. American Psychologist 37:l22-l47. 94

Becker, J. H. l982. Microcomputers in the Classroom: Dreams and Re- alities. Report No. 3l9. Baltimore, Md.: Johns Hopkins Univer- sity Center for Social Organization. Benbow, C. P., and J. C. Stanley. l980. Sex differences in mathematical ability: Fact or artifact? Science 2l0:l262-l264. Benbow, C. P., and J. C. Stanley. l983. Sex differences in mathematical ability: More facts. Science 222:l029-l03l. Bogart, K. l984. Toward Equity: An Action Manual for Women in Academe. Washington, D.C.: Association of American Colleges. Boss, J. A. l982. Sexism among the micros. Computing Teacher 9:55-57. Brown, L. l984. Retention of Women in Science Disciplines: Results of the First Year. West Lafayette, Ind.: Purdue University. Bureau of the Census. l985. Statistical Abstract of the United States: l986 (l06th ed.). Washington, D.C.: U.S. Department of Commerce. Bureau of Labor Statistics. l986. Occupational Projections and Train- ing Data. Washington, D.C.: U.S. Department of Labor. Byrne, E. M. l985. Women and Engineering: A Comparative Overview of New Initiatives. Bureau of Labour Market Research Monograph Series No. ll. Sydney: Australian Government Publishing Service. Campbell, P. F., and G. P. McCabe. l984. Predicting the success of freshmen in a computer science major. Communications of the ACM 27(ll):ll08-lll3. Chronicle of Higher Education. l984 Fall Enrollment. January 22, l986, p. 26. Cole, J. R. l979. Fair Science, Free Press. New York: Macmillan. Connelly, T., and A. L. Porter. l978. Women in Engineering: Policy Recommendations for Recruitment and Retention in Undergraduate Programs. Atlanta: Georgia Institute of Technology. Corbett, J. G., S. Estler, W. Johnston, M. D. Ott, H. Robinson, and G. R. Sell. l980. Women in Engineering: An Exploratory Study of Enrollment Factors in the Seventies. Boulder, Colo.: National Center for Higher Education Management Systems. Davis, J. A. l965. Undergraduate Career Decisions. Chicago: Aldine Press. Educational Testing Service. l97l-l985. National College-Bound Seniors. An annual report. Princeton, N.J.: Admissions Testing Program, The College Board. Engineering Manpower Commission. l979. Engineering and Technology En- rollments, Fall l979, Part I: Engineering. New York: American Association of Engineering Societies. Engineering Manpower Commission. l984. Engineering and Technology En- rollments, Fall l984, Part I: Engineering. New York: American Association of Engineering Societies. Fetler, M. l985. Sex differences on the California statewide assess- ment of computer literacy. Sex Roles l3:l8l-l9l. Hawkins, J. l985. Computers and girls: Rethinking the issues. Sex Roles l3:l65-l79. Hess, R. D., and I. T. Miura. l985. Gender differences in enrollment in computer camps and classes. Sex Roles l3:l93-203. Humphreys, S. M., ed. l982. Women and Minorities in Science: Strate- gies for Increasing Participation. Washington, D.C.: American Association for the Advancement of Science. 95

Ranter, R. M. l977. Men and Women of the Corporation. New York: Basic Books. Lantz, A., L. Elliott, and L. Pox. l979. Critical Mass and Social Sup- port for Women Choosing Nontraditional Careers. Washington, D.C.: National Institute for Education. Lantz, A., A. S. West, and L. Elliott. l976. An Impact Analysis of Sponsored Projects to Increase the Participation of Women in Careers in Science and Technology. University of Denver Interim Technical Report. Denver: Denver Research Institute. LeBold, W. K., B. R. Butler, and E. M. Knigga. l983. Putting It All Together. West Lafayette, Ind.: Purdue University. Lepper, M. R. l985. Microcomputers in education: Motivational and social issues. American Psychologist 40:l-l8. Linden, K. W., C. M. Jagacinski, W. K. LeBold, and K. D. Shell. l985. Predicting persistence in engineering for undergraduate women. In the Third International Girls in Science and Technology Conference. London: Centre for Science and Mathematics Education. Linn, M. C. l985. Fostering equitable consequences from computer learning environments. Sex Roles l3:229-239. Linn, M. C., and J. Dalbey. l985. Cognitive consequences of program- ming instruction; instruction, access, and ability. Educational Psychologist 20:l9l-206. Lockhard, J. l980. Computers blossom at a small school in Iowa. In- structional Innovator 28:l4-25. Lockheed, M., A. Nielsen, and M. Stone. l983. Sex differences in mi- crocomputer literacy. Paper presented at the National Educational Computer Conference, Baltimore, Md., August l983. Luchins, E. H., and A. S. Luchins. l98l. Letters. Science 2l2:ll7- ll8. Maccoby, E. E. l966. Sex differences in intellectual functioning. In the Development of Sex Differences, E. E. Maccoby, ed. Stanford, Calif.: Stanford University. Mandinach, E., and C. W. Fisher. l985. Individual Differences and Acquisition of Computer Programming Skill. Berkeley: University of California. Miura, I. T. l984. Processes Contributing to Individual Differences in Computer Literacy. Ph.D. dissertation, Stanford University. Miura, I. T., and R. D. Hess. l984. Sex differences in computer ac- cess, interest, and usage. Paper presented at the Annual Conven- tion of the American Psychological Association, Anaheim, Calif., June l984. Miura, I. T., and J. A. Miura. l984. Computer self-efficacy: A con- tributing factor to gender differences in the election of computer science courses. Paper presented at the Fifth International Con- gress of Mathematical Education, Adelaide, Australia, June l984. National Center for Education Statistics. l97l-l983. Earned Degrees Conferred series. Washington, D.C.: U.S. Office of Education. National Center for Education Statistics. l982. Digest of Education Statistics. Washington, D.C.: U.S. Department of Education. National Research Council, Commission on Engineering and Technical Sys- tems. l985a. Engineering Education and Practice in the United States. Washington, D.C.: National Academy Press. 96

National Research Council, Commission on Engineering and Technical Sys- tems. l985b. Engineering Graduate Education and Research. Wash- ington, D.C.: National Academy Press. National Research Council, Commission on Engineering and Technical Sys- tems. l985c. Engineering Infrastructure Diagramming and Model- ing. Washington, D.C.: National Academy Press. National Research Council, Office of Scientific and Engineering Person- nel. l973-l986. Doctorate Recipients from United States Univer- sities: Summary Report. An annual report. Washington, D.C.: National Academy Press. National Science Board, Commission on Precollege Education in Mathe- matics, Science and Technology. l984. Educating Americans for the 2lst Century. Report No. CPCE-NSF-04. Washington, D.C.: U.S. Government Printing Office. NSF (National Science Foundation). l983. Academic Science/Engineer- ing: Graduate Enrollment and Support. Washington, D.C.: U.S. Gov- ernment Printing Office. NSF. l986a. Academic Science/Engineering: Graduate Enrollment and Support. Washington, D.C.: U.S. Government Printing Office. NSF. l986b. Women and Minorities in Science and Engineering. Wash- ington, D.C.: U.S. Government Printing Office. Ott, M. D. l978. Retention of men and women engineering students. Research in Higher Education 9:l37-l50. Pallas, A. M., and K. L. Alexander. l983. Sex differences in quanti- tative SAT performance: New evidence on the differential course work hypothesis. American Educational Research Journal 20:l65-l82. Pea, R. D., J. Hawkins, C. A. Clement, and R. Maby. l984. The Devel- opment of Expertise in Logo by Children. New York: Bank Street College of Education, Center for Children and Technology. Pister, K. S., and S. M. Humphreys. l986. Affirmative Action Report. Berkeley: Office of Women and Minority Graduate Student Matters, College of Engineering, University of California. Saunders, J. l979. What are the real problems involved in getting computers into the high school? Mathematics Teacher (May):443- 447. Schubert, J. G. l984. Females and microcomputers in schools: Some insights into traditional patterns. Paper presented at the annual meeting of the Special Interest Group, Research on Women in Educa- tion, American Educational Research Association, Los Angeles, April l984. Sheingold, K., J. Kane, and M. Endreweit. l983. Microcomputer use in schools: Developing a research agenda. Harvard Educational Re- view 53:4l2-432. Shell, K. D., W. K. LeBold, K. W. Linden, and C. M. Jagacinski. l985. Retention research in engineering education. In Proceedings of the l985 Frontiers in Education Conference. Arlington, Texas: American Society for Engineering Education. Stage, E. K., and R. Karplus. l98l. Letters. Science 2l2:ll4. Stein, A., and J. Smithells. l969. Age and sex differences in chil- dren's sex-role standards about achievement. Developmental Psy- chology l:252-259. 97

Tomizuka, C., and S. Tobias. l98l. Letters. Science 2l2:ll4. Turkic, S. l984. The Second Self: The Computer and the Human Spirit. New York: Simon and Schuster. U.S. Department of Education. l982. Surveys of High School and Beyond, l980. Washington, D.C.: U.S. Government Printing Office. Vetter, B., and E. Babco, eds. l986a. Professional Women and Minori- ties. An annual publication. Washington, D.C.: Commission on Pro- fessionals in Science and Technology. Vetter, B., and E. Babco, eds. l986b. Manpower Comments. A monthly publication. Washington, D.C.: Commission on Professionals in Sci- ence and Technology. Winkle, L. W., and W. M. Mathews. l982. Computer equity comes of age. Phi Delta Kappan 63:3l4-3l5. 98