Science, Computers, and Gender Equity

Is there a role for single-sex education?

Leonard Sax, M.D., Ph.D.

Title IX of the Education Amendments of 1972 prohibits discrimination on the basis of sex. Before passage of Title IX, school districts often made a practice of tracking girls and boys into different courses: woodworking and mechanical drawing for boys, home economics and shorthand for girls. Enrollment in upper-level math and science courses was overwhelmingly male when Title IX became law (Petty & Green, 1977).

More than 30 years have passed since enactment of Title IX in June 1972. Have girls made inroads into traditionally "male" subjects such as physics and computer science over the past 30 years? The National Council for Research on Women recently released a study showing that although 56 percent of high school students taking Advanced Placement tests are girls, only 22 percent of students taking the physics AP exam are girls, and only 10 percent of students taking the computer science AP exam are girls (NCRW, 2001). An earlier investigation likewise found that enrollment in high school courses involving computers has become more male-dominated in recent years (Littleton & Bannert, 1999).

The goal of achieving gender equity in math and computer science has not been met. The trend may even be headed in the wrong direction. What can be done?

There is substantial evidence that girls' interest in math and computer science is stimulated more effectively in single-sex educational settings (e.g. Culley, 1993; Newton & Beck, 1993). In a moment we will consider some of the evidence for and against the hypothesis that single-sex education can promote gender equity, particularly in subjects such as physics and computer science.

Until very recently, "single-sex education in public schools" would have sounded oxymoronic to most American educators. The regulations implementing Title IX were widely understood as prohibiting single-sex education in public schools (Sax, 2001). But on May 8, 2002, the Department of Education announced that it was beginning the process of rewriting the regulations implementing Title IX, in order "to provide more flexibility for educators to establish single-sex classes and schools at the elementary and secondary levels" (U. S. Department of Education, 2002, p. 31098). This Departmental action was mandated by a clause in the No Child Left Behind Act of 2001 authorizing federal funding for single-sex educational programs (see §5131(a)(23) and §5131(c) of the Act). The 60-day period for public comment ended on July 8, 2002.

A few school districts are not even waiting to see the final regulations. In May and June 2002, school districts in Philadelphia and Louisville announced that they were converting existing coed schools to single-sex academies, effective in September 2002 (Snyder, 2002; Kenning, 2002). At the same time, the Houston school district announced the establishment of a new single-sex charter school, opening in September 2002 (Markley, 2002).

Can single-sex schools and/or classrooms help to achieve gender equity for girls in math and computer science? Or do they create more problems than they solve?

As British researcher Eileen Scanlon has observed, "there is a growing realization among science educators that particular features of the learning experience are altered by the gender composition of the groups" (Scanlon, 2000, p. 463). Jean and Geoffrey Underwood published work in 1990 showing that students in girl-boy pairs performed significantly worse on a computer-based language task than did girl-girl or boy-boy pairs (Underwood, McCaffrey and Underwood, 1990). A similar result was reported in a study employing a computer-based interactive video task (Dalton, 1990). Pozzi, Healy and Hoyles (1993) suggested that when care was taken to ensure that the task was gender-neutral, no gender effects were seen in performance scores. However, they acknowledged that different gender groupings "were associated with different forms of group processes" (p. 222). In particular, they observed that cross-gender conflicts could give rise to a fragmented work style.

Subsequent work by the Underwoods (Underwood & Underwood, 1997) clarified features of the mixed-sex dyad which may help to explain the poor performance of mixed-sex groups, at least in younger children. The Underwoods gave 31 pairs of 8-year-olds a computer-based language task. Children were randomly assigned either to girl-girl, girl-boy, or boy-boy pairs. Each child within a pair was matched with the other for reading ability. The Underwoods found a dramatic difference in story recall, depending on the gender composition of the pair. Boys in boy-boy pairs performed least well, while girls assigned to girl-girl pairs obtained the highest scores. The most striking finding, however, was that girls in girl-boy pairs performed almost as badly as the boys did. Just putting a girl with a boy degraded her performance by roughly 50% on this computer-based task. This effect was highly significant (p < 0.001). However, other studies found no impairment of girls' performance in mixed-sex environments (e.g. Light, Littleton, Messer, & Joiner, 1994).

The Underwoods videotaped all the children during the task. Noting that "the boys in mixed pairs tended to take control of the task," they suggested that "this imbalance in control may have contributed to [the girls'] diminished enthusiasm for the task," which might explain the poor recall of girls in the mixed pairs. If this explanation is valid, then the negative effect of mixed-sex groupings (on girls) might be minimized if each girl were given her own computer.

A remarkable study by Paul Light, Karen Littleton and their associates tested this specific hypothesis. Light et al. (2000) gave 11- and 12-year-old girls and boys each their own computer. Some girls were in girl-only groups; others were in groups with both girls and boys. "The children were simply working alongside one another, on separate machines and without any overt interaction," the authors emphasize (p. 488). Nevertheless, girls' performance deteriorated by roughly 40% in the mixed-sex grouping compared with the all-girls grouping. The authors concluded that this "in-between condition -- probably the usual one in most classrooms -- where boys and girls work in close proximity but with minimal interaction, may represent the worst of all worlds" (p. 495).

A recent year-long study from Germany (Hoffmann, 2002) suggests that a similar phenomenon may take place in physics classrooms. In this study, students in six different coeducational schools were assigned either to have all their physics classes in coed classes, or to have half their physics instruction in coed classes and half in single-sex classes. (No student had the opportunity to be in a class which was always single-sex.) Girls who had half their classes in single-sex classes did significantly better than girls who were always in coed classes (p < 0.05); and they remained just as interested in physics at the end of the school year as they had been at the beginning, whereas girls who had only coed classes were significantly less interested in physics at the end of the year than they had been at the beginning.

Large-scale studies have demonstrated the beneficial effects of single-sex educational settings for girls, particularly in math and science. Hamilton (1985) surveyed high school students throughout Jamaica, a country in which single-sex public education remains widely available. She was able to gather test data for almost one-seventh of the nation's entire population of high school students. In her survey, girls at single-sex high schools did significantly better in math and science than girls at coed high schools. In fact, the girls at single-sex high schools sometimes outperformed the boys in math and science.

Hamilton (1985) concluded that "the effects of sex stereotyping are more sharply apparent in the coeducational setting, for here there appears to be a stronger need to differentiate between the sexes. Thus, even if subjects such as mathematics and the sciences are technically available to girls in this type of school, they are more likely than their single-sex institutional counterparts to get the message that such subjects are unfeminine, and beyond their grasp intellectually." (p. 545)

However, Hamilton's study did not control for prior attainment. It is possible that girls with high ability in math and science, and/or with high prior achievement in math and science, were more likely to attend single-sex schools where such schools were available. If that were the case, then the superior performance of girls at single-sex schools might have been a result of the type of girls attending those schools, rather than the single-sex character of the school itself.

A more recent study of more than 45,000 students from over 400 schools in Hong Kong, where about one-tenth of the public schools are single-sex, reported similar results and drew similar conclusions (Wong, Lam, & Ho, 2000). These investigators attempted to control for pre-existing academic attainment by normalizing scores on the test (given in 11th grade) based on each individual student's scores prior to enrolling in the high school. Even after controlling for pre-existing ability, these investigators found that girls in single-sex schools did significantly better on the mathematics examination than did girls in coed schools (separate scores were provided only for English, Chinese, and mathematics, not for the sciences; see table 3 from their report).

The largest study comparing the performance of students at single-sex and coeducational schools was recently released by the National Foundation for Educational Research (NFER), a British non-profit organization. The NFER study (Spielhofer, O'Donnell, Benton, et al., 2002) examined the test results of 369,341 pupils from 2,954 schools in the United Kingdom. Each student's results on standardized tests in 11th grade were adjusted for school-level and pupil-level factors, including each student's pre-existing ability as measured by previous test performance. Even after such adjustments, the authors found that "single-sex schooling has a small but significant effect on girls' achievement, with an especially marked effect on their performance in science. . . this was the largest measured effect of single-sex schooling on girls' achievement and was, more or less, constant across the ability range." (Spielhofer et al., 2002, p. 32) They also found girls in girls-only schools were 30 to 40% more likely to take advanced science courses (analogous to our AP courses). In this respect, "being in a girls' school counteracted the effect of being a girl, since girls in mixed schools [i.e. coeducational schools] were less likely to take [advanced] sciences than boys. . ." (p. 42). More generally, they found that girls' schools are "helping to counter rather than reinforce the distinctions" between "girls' subjects" such as English and foreign languages and "boys' subjects" such as physics and computer science (p. 43).

Critics Respond

Critics of single-sex education commonly advance one of two arguments. The first argument is simply that adult women and men will be able to work together better, and relate to one another better, if they have had the experience of going to school together as children. This hypothesis is intrinsically difficult to test. However, the available evidence -- particularly in regard to interpersonal relationships -- does not support the detractors' hypothesis. Katherine Sanders and Neville Bruce recently examined the frequency and duration of heterosexual romantic episodes in 17-, 18-, and 19-year-olds in Australia, a country in which single-sex public schools are widely available (Bruce & Sanders, 2001). They found that students from single-sex schools were no more and no less likely to be involved in romantic episodes than students from coed schools, nor did the average duration of the episode differ as a function of school type (9 weeks if the romantic interest was never reciprocated, 12 weeks if it was reciprocated). These authors acknowledged that when they launched the study, they had "expected that single-sex schooling might inhibit incidence of romances, at least during the school years. But this view was supported by only two of eight possible comparisons and the trend in four [of eight] possible comparisons was in the opposite direction, although not significantly so. Thus, it would seem that students from single-sex schools are not noticeably thwarted by any lack of opportunity or experience in the single-sex school system from experiencing romantic episodes, either at school or later in their early university years" (p. 183).

Cornelius Riordan published an important longitudinal study examining the experience of students 7 and 14 years after graduation from either single-sex or coeducational high schools. Riordan found no significant differences in the likelihood of getting married or getting divorced between married graduates of single-sex and coed schools. On a 42-point scale of "marital happiness," there was again no difference between graduates of single-sex schools and coed schools when assessed either 7 or 14 years after graduation from high school. Seven years after graduation, both women and men who had graduated from single-sex high schools reported significantly higher educational attainment than graduates of coed schools; 14 years after graduation, the difference was no longer statistically significant -- perhaps owing in part to attrition and a smaller sample size -- but the trend still favored graduates of single-sex schools (Riordan, 1990, especially table 6.6, p. 130).

The second argument against single-sex education derives from the rhetoric of the civil rights movement. This criticism rests on making an analogy between race and sex. On August 1, 1994, when Senator John Danforth (R-Missouri) proposed allowing single-sex education in public schools, Senator Edward Kennedy (D-Massachusetts) insisted that he would not "support taxpayers' money to be utilized in ways that are going to further segregation, whether it is going to be on [the basis of] race or religion or ethnicity or gender . . . How long is it going to take before the next amendment says, ‘Well, all right, let us separate the blacks out, let us put Asians [over here], let us put the women students [over there], let us put the browns out here.' . . . And this could lead to ‘Maybe we can experiment with single-sex programs in housing, as well as employment and education. Just let women in only certain parts of the house. Maybe this would reduce some of the violence. Or, if not the women, maybe we can find a way to just put the males together in certain kinds of housing. Maybe that's the way we can reduce violence'" (U. S. Government Printing Office, 1994, pp. 18744 and 18746). Advocates of this position are not impressed by evidence that girls do better in single-sex classes, any more than they would be swayed, for example, if they were to be shown evidence that Black students do better in all-Black classes. Racial segregation is of course not acceptable to anyone. These critics regard sex and race as comparable categories; therefore segregation by sex is also unacceptable to them.

This argument was repeated by opponents of single-sex education after the Department of Education's May 2002 proposal to allow single-sex education in public schools. "We believe that separate is never equal," said Nancy Zirkin in May 2002 (Zirkin is director of public policy for the American Association of University Women). "If we were talking about this in a race context do you think we would be talking about it? I don't think so" (Povich, 2002). Nationally-syndicated columnist Ellen Goodman made a similar argument. "How did we get here?" she asked. "A generation ago, in court and Congress, we agreed that separate was not equal in public schools, whether we were talking about race or gender . . . What [government officials] really want is to revise federal regulations to allow funding for, well, resegregation." (Goodman, 2002)

Arguments such as these, equating race and sex, carried more weight 30 years ago, before the advent of MRI scans and quantitative histological techniques demonstrating innate sex differences in the brain. Thirty years ago, few sex differences in the human brain were known. In 1969, for example, Naomi Weisstein wrote a provocative essay entitled "Woman as Nigger," which included this memorable remark: "Except for their genitals, I don't know what immutable differences exist between men and women" (Weisstein, 1969). In that era, many progressive thinkers assumed that there were no innate, genetically-programmed differences between a girl's brain and a boy's brain. Beginning from that assumption, some scholars argued that observed sex differences in how children learn and behave must be socially constructed (e.g. Kessler & McKenna, 1978). From this perspective, any evidence that girls do better in single-sex schools would merely be proof of the destructive effects of patriarchal culture in coeducational schools. Peter Glick and Susan Fiske recently advanced this argument: "In other words, sex differences in personality are to some degree caused by structural sexism in societies . . . Thus, the relationship between a sexist society and observed sex differences is circular" (Glick & Fiske, 2002, p. 445).

The introduction of new technologies for examining the human brain has deprived such arguments of much of their force. Innate differences between female and male brains are much larger than previously thought, with little overlap in morphology between the sexes (although these findings are not yet well known within the professional education community). For example, scientists are now able to distinguish whether a pregnant woman is carrying a female baby or a male baby, just on the appearance of the baby's brain during an ultrasound examination (Achiron, Lipitz, & Achiron, 2001). In another recent study, the microscopic structure of the three-week-old girl's brain has been shown to be dramatically different from the microscopic structure of the brain of the three-week-old boy. In the girl's brain, connections between brain cells are much richer and denser (Cordero, Valenzuela, Torres, & Rodriguez, 2000). In this study, sex differences in brain anatomy are readily apparent to the naked eye.

Biologically-programmed differences in brain structure and function persist throughout childhood and into adulthood. Sandra Witelson (1995) examined brain tissue from language association areas under the microscope. She found that brain cells are packed much more densely in the female brain. In Witelson's study, the brain tissue of every single female had a higher concentration of cells than any male had. These studies, and many others like them, clearly demonstrate that there are "fundamental gender differences in the structure of the human cerebral cortex" (Rabinowicz et al., 2002, p. 52). No such racial differences exist. You can't tell by looking at a slice of brain tissue whether you are looking at tissue from a Black person or a White person or a Hispanic person. But you can tell -- often quite easily -- whether you are looking at tissue from a girl or a boy.

There is growing evidence that girls and boys learn differently because their brains are built differently (e.g. NASSPE, 2002). These differences in brain structure are not socially constructed, but are genetically programmed and present at birth. As we saw, the second argument against single-sex education rests on the analogy between race and sex. This analogy appears not to be sustainable in the light of modern scientific evidence. Racial differences in learning are surely socially constructed. Sex differences in how children learn appear, at least in part, to be biologically programmed.

Directions for further research

Although single-sex education holds promise as a means of achieving gender equity for girls in traditionally male-dominated subjects, many questions remain unanswered. For example: do single-sex schools have advantages or disadvantages compared to single-sex classrooms located within coeducational schools? What is the effect of single-sex education with respect to gender equity for boys -- for example, in traditionally female-dominated subjects such as foreign languages, or dance? And, is there any particular age for which single-sex education is particularly beneficial? With the expected liberalization in federal regulations governing single-sex education in public schools, answers to these questions may be forthcoming.



©2002 Leonard Sax


If you would like to be on our mailing list, and receive quarterly updates on what's going on with single-sex public education in the United States, or if you just would like to share your thoughts or ideas about this topic with us, you can write NASSPE at Post Office Box 108, 19710 Fisher Avenue, Poolesville, Maryland, 20837. Or, you can send us an e-mail, or call us at 301 972 7600, or send a fax to 301 972 8006.


Click here to return to the NASSPE home page.


References

Achiron, R., Lipitz, S., & Achiron, A.. (2001) Sex-related differences in the development of the human fetal corpus callosum: in utero ultrasonographic study. Prenatal Diagnosis, 21:116-120.

Bruce, N., & Sanders, K. (2001). Incidence and duration of romantic attraction in students progressing from secondary to tertiary education. Journal of Biosocial Sciences, 33:173-184.

Cordero, M. E., Valenzuela, C. Y., Torres, R., & Rodriguez, A. (2000). Sexual dimorphism in number and proportion of neurons in the human median raphe nucleus. Developmental Brain Research, 124:43-52.

Culley, L. (1993). Gender equity and computing in secondary schools: issues and strategies for teachers. In J. Beynon & H. Mackay, Computers into classrooms: more questions than answers (pp. 147-159), London: Falmer Press.

Dalton, D. W. (1990) Effects of cooperative learning strategies on achievement and attitudes during interactive video. Journal of Computer-Based Instruction, 17(1):8-16.

Glick, P., & Fiske, S. T. (2002). Ambivalent responses. American Psychologist, 57(6):444-446.

Goodman, E. (2002). Back to single-sex schools? Boston Globe, May 16, p. A17.

Hamilton, H. (1985). Performance levels in science and other subjects for Jamaican adolescents attending single-sex and coeducational high schools. International Science Education, 69(4):535-547.

Hoffmann, L. (2002). Promoting girls' interest and achievement in physics classes for beginners. Learning and Instruction, 12:447-465.

Jimenez, E., & Lockheed, M. E. Enhancing girls' learning through single-sex education. Educational Evaluation and Policy Analysis, 11:117-142.

Kenning, C. (2002). Southern Middle will separate boys, girls. Courier-Journal (Louisville, Kentucky), June 18.

Kessler, S., & McKenna, W. (1978). Gender: an ethnomethodological approach, Chicago: University of Chicago Press.

Lee, V. E., & Lockheed, M. E. (1990). The effect of single-sex schooling and attitudes in Nigeria. Comparative Education Review, 34:209-231.

Light, P., Littleton, K., Bale, S., Joiner, R., & Messer, D. (2000). Gender and social comparison effects in computer-based problem solving. Learning and Instruction 10:483-496.

Light, P., Littleton, K., Messer, D., & Joiner, R. (1994). Social and communicative processes in computer-based problem solving. European Journal of Psychology of Education, 9(1):93-109.

Littleton, K., & Bannert, M. (1999). Situating differences: the case of gender and computer technology. In J. Bliss, R. Saljo, & P. Light, Learning sites: social and technological resources for learning (pp. 171-182). New York: Pergamon.

Markley, M. (2002). Helping boys make the grade: new law on single-sex schools opens way for Houston project. Houston Chronicle, May 2, p. A1.

NASSPE (2002). Are there significant differences between a girl's brain and a boy's brain? National Association for Single-Sex Public Education. Online here.

Newton, P., & Beck, E. (1993). Computing: an ideal occupation for women? In J. Beynon & H. Mackay, Computers into classrooms: more questions than answers (pp. 130-146), London: Falmer Press.

NCRW (2001). Balancing the Equation: where are women and girls in science, engineering and technology? National Council for Research on Women monograph.

Petty, J., & Green, W. (1977). Almost As Fairly: the first year of Title IX implementation in six southern states. Atlanta, Georgia: Southeastern Public Education Program.

Pozzi, S., Healy, L., & Hoyles, C. (1993). Learning and interaction in groups with computers: when do ability and gender matter? Social Development, 2(3):222-241.

Povich, E. S. (2002). Clinton for same-sex schools. Newsday (New York), May 10, p. A4.

Rabinowicz, T., MacDonald-Comber Petetot, J., Gartside, P., Sheyn, D., Sheyn, T., & De Courten-Myers, G. (2002). Structure of the cerebral cortex in men and women. Journal of Neuropathology and Experimental Neurology, 61(1):46-57.

Riordan, C. (1990). Girls and Boys in School: together or separate? New York: Teachers College Press.

Sax, L. (2001). Rethinking Title IX. Washington Times, July 2, p. A17.

Scanlon, E. (2000). How gender influences learners working collaboratively with science simulations. Learning and Instruction, 10:463-481. (Scanlon writes in British English: I have substituted "realization" for "realisation" and "among" for "amongst.")

Snyder, S. (2002). A school trial will separate the sexes. Philadelphia Inquirer, June 17, p. A1.

Spielhofer T., O'Donnell, L., Benton, T., Schagen, S., & Schagen, I. (2002). The impact of school size and single-sex education on performance. London (UK): National Foundation for Educational Research, July 2002.

Underwood, G., McCaffrey, M., & Underwood, J. (1990) Gender differences in a co-operative computer based language task. Educational Researcher 32(1):44-49.

Underwood, G., & Underwood, J. (1997). Children's interactions and learning outcomes with interactive books. Paper presented at the CAL (Computer Assisted Learning) Conference, April 2 1997, at the University of Exeter.

U. S. Department of Education. (2002). Proposed Rule (Notice of Intent to Regulate). Federal Register, 67(89):31098-31099, May 8.

U. S. Government Printing Office. (1994). Congressional Record, 140(13):18744-18746.

Weisstein, N. (1969). "Woman as Nigger," Psychology Today, October 1969.

Witelson, S., Glezer, I., & Kigar, D. (2001) Women have greater density of neurons in posterior temporal cortex. Journal of Neuroscience, 15:3418-3428.

Wong, K., Lam, Y., & Ho, L. (2001) The effects of schooling on gender differences. Paper presented November 30 2001 at a conference sponsored by the Equal Opportunities Commission (Hong Kong), "Boys and Girls in the 21st Century." Document available online http://www.eoc.org.hk/_File/conference/BG_21/01/GenderDiff_China.doc



©2002 Leonard Sax


If you would like to be on our mailing list, and receive quarterly updates on what's going on with single-sex public education in the United States, or if you just would like to share your thoughts or ideas about this topic with us, you can write NASSPE at Post Office Box 108, 19710 Fisher Avenue, Poolesville, Maryland, 20837. Or, you can send us an e-mail, or call us at 301 972 7600, or send a fax to 301 972 8006.


Click here to return to the NASSPE home page.