Appendix to Elizabeth Anderson’s Review of Scrutinizing Feminist Epistemology (Pinnick, Koertge, and Almeder 2003): Christina Hoff Sommers’ “Where the Boys Are”

Christina Hoff Sommers’ chapter, “Where the Boys Are” has nothing to say about feminist epistemology, feminist philosophy of science, or feminist science studies. It engages no philosophical issues. Instead, it is an attack on a few pieces of feminist research that claim to find gender differences in mental health and gender inequity in the schools. Presumably it is intended as a case study in how feminist-inspired research is done, attempting to demonstrate that it is false and politically biased.

Sommers argues that feminist researchers--in particular, Carol Gilligan, Mary Pipher, David and Myra Sadker, and researchers for the American Association of University Women (AAUW)--have “manufactured” a crisis concerning girls (124). The feminist researchers claim that adolescent girls suffer from lower self-esteem and mental health relative to boys, and that schools shortchange girls in a variety of ways. Sommers argues that the truth is the reverse: in fact it is boys who are suffering from neglect by the schools and a crisis in educational achievement. She claims that the feminist education researchers paint a misleading picture by ignoring boys’ difficulties in school. But Sommers’ own work misrepresents the research she criticizes.[1] Let us count the ways.

Misleading suggestions of data suppression. Sommers suggests that the feminist researchers she criticizes have suppressed their data and inconvenient findings. For example, she insinuates that Myra and David Sadker have mysteriously failed to publish or make available their study that found that boys called out unsolicited answers eight times more than girls (107, 126n5). In fact, the Sadkers have acknowledged that the 8-to-1 callout ratio, based on preliminary findings, was erroneous. They have published a more comprehensive study superceding their symposium presentation, which found a lower callout ratio still favoring boys--consistent with the 2-to1 ratio reported by Sommers’ own source (107). This study confirmed the far more important finding of their preliminary report, that teachers tend to reward boys’ callouts with positive interactions, but discourage girls’ callouts by correcting their conduct (Sadker and Sadker 1984, 114). By failing to report that the Sadkers have made public available their final, complete, and corrected study, Sommers misleadingly suggests that they are trying to hide something.

Sommers accuses the AAUW of failing to publish data that show that students perceive that teachers favor girls over boys. Yet, in reporting this data herself, she cites an AAUW publication (124)! Although she falsely labels the data “unpublished,” what she really means is that the AAUW did not report this data in their executive summary of their published study. Sommers thinks this is damning, because she accepts students’ perceptions of antiboy bias at face value. She is mistaken. Although students perceive that teachers compliment, pay more attention to, and call on girls more often than boys, objective measures of teacher interaction, reported in numerous studies, reveal that teachers tend to favor boys on these criteria (Sadker, Sadker, and Klein 1991, 294–304). If even the relatively little attention girls receive from teachers is seen as favoring girls over boys, this is evidence, not of a classroom climate favoring girls, but of sexism on the part of students.[2]

Sommers claims that the AAUW failed to publicize a study it commissioned from Valerie Lee (Lee, Chen, and Smerdon 1996), because it supposedly contradicted their earlier, highly publicized study that found the schools shortchange girls (117). Lee herself has publicly repudiated this charge, observing that she did not authorize the AAUW to publicize her study because she and her fellow researchers retained the copyright, so they could publish their findings in peer-reviewed journals (Lee 1996). Sommers’ contribution to SFE reprints work she published in 2000. Thus, in SFE she makes an unfounded, malicious accusation for the second time, although she had years to correct herself.

Refuting apples with oranges. Sommers tries to discredit claims of school bias against girls by selectively citing certain aggregate statistics on education outcomes in which boys do less well than girls. But she mostly ignores the outcomes that favor boys, which were the focus of the feminist research: for example, the fact that boys tend to do better on most high-stakes standardized tests, that they dominate the physical sciences, engineering, and computer science, and that girls are concentrated in vocational programs, college majors and graduate programs that lead to lower-paying, lower-prestige careers than their male counterparts (AAUW 1992, 24, 27-8, 42-3, Sadker 1996). Furthermore, many complaints about gender bias in schools focus not on the education outcomes Sommers stresses, but on various education inputs and other factors affecting the school experience. The famous AAUW (1992) report, How Schools Shortchange Girls, which is the focus of Sommers’ critique, complains about the paucity of female school principals and superintendents, the failure of many schools to understand or implement Title IX requirements concerning educational programs for pregnant teens and teen mothers and gender equity in the provision of sports opportunities, bias in the curriculum (the invisibility of women, promotion of sex stereotypes, lack of multicultural content, etc.), sex bias in the classroom (e.g., patterns of student-teacher interaction favoring boys), sexual harrassment of girls and homophobic harrassment of students perceived to be gay, and sex education curricula that reinforce girls’ shame about and boys’ ignorance of menstruation (AAUW 1992, 7, 8, 37-41, 45, 61-7, 68-71, 73-4, 77). Sommers mostly ignores these inequities.

These inequities cannot be refuted or dismissed by citing aggregate statistics on education outcomes, such as school grades and graduation rates, that favor girls. Such statistics cannot rule out the possibility that these inequities do depress girls’ academic outcomes. In any event, the importance of these inequities is not limited to their impact on girls’ aggregate academic outcomes. Sexual harassment is bad in itself. Sex bias in the curriculum is objectionable for reinforcing sexist stereotypes. Schools’ failure to provide educational opportunities for pregnant teens and teen mothers cannot be excused by pointing out that girls overall have higher graduation rates than boys. Pace Sommers (107), the bias toward boys in classroom interaction is objectionable even if it doesn’t depress girls’ academic outcomes, because it expresses and reinforces sexist norms that treat boys’ contributions as more important than girls’, and entitle boys to break rules of polite conversational interaction that are enforced against girls. Sommers cannot refute the claim that schools shortchange girls in respects A, B, and C by pointing out that they do not shortchange girls in respects X, Y, and Z.

Falsely claiming that feminist research ignores respects in which boys are disadvantaged. Sommers accuses the AAUW of ignoring and dismissing boys’ problems (123). To support these accusations, Sommers cites research by Lee, Chen, and Smerdon (1996) (sponsored by the AAUW itself!), Kleinfeld (1998) and others on gender differences in educational outcomes that she claims contradicts what she represents as a monolithic picture of antigirl bias in the schools painted by the AAUW in its 1992 report. But the idea that the 1992 AAUW report painted a monolithic picture of antigirl bias is absurd. In fact, the 1992 AAUW report scrupulously noted many of the gender differences in outcomes disfavoring boys that Sommers implies were only revealed later by her favored researchers. (Some of the others were revealed only in research that postdated the 1992 AAUW report.[3]) Check it out the parallel claims for yourself:

Girls attempt suicide more than boys, but boys more often succeed in killing themselves (110). (AAUW 1992, 79)

Boys are more often assigned to special education instruction than girls (109). (AAUW 1992, 19).

Girls get higher grades than boys (108). (AAUW 1992, 22, 33, 52)

More boys than girls are held back (109). (AAUW 1992, 35)

Gender differences in enrollment in high school math and biology courses are small (110). (AAUW 1992, 26-27)

Boys have higher high school dropout rates than girls. (AAUW 1992, 47)

More girls than boys take AP exams (108). (AAUW 1992, 36)

The gender gap in math achievement tests (favoring boys) is small and narrowing (116). (AAUW 1992, 24)

Boys have consistently scored lower than girls on the NAEP reading and writing tests (116). (AAUW 1992, 23)

The pattern of gender differences in school achievement is inconsistent in direction, sometimes favoring boys, sometimes girls (117). (AAUW 1992, 22-27, 33)

Girls have higher educational aspirations than boys (108, 119). (AAUW 1992, 35)

Girls have higher college attendance rates than boys (113-114). (AAUW 1992, 47, 52)

African-American women attend college at far higher rates than African-American men (122). (AAUW 1992, 47-8)

Sommers claims that studies documenting gender gaps in educational achievement that disfavored boys “began to surface” only in the late 1990s (122). In reality, as the above citations demonstrate, the 1992 AAUW report candidly reported numerous gender gaps disfavoring boys. Sommers could have written her article about the crisis in boys’ achievement by heavily relying on this report. But that would have ruined her story, according to which the 1992 AAUW report promulgated an “antiboy climate” (122) that denied that schools were failing boys, too.

Omitting contextual facts that make sense of feminist concerns. Sommers implies that, with so many aggregate educational outcomes favoring girls, feminists are unjustified in worrying about girls’ achievement in respects in which they outperform boys. But Sommers ignores other data cited by feminist researchers that justify their concerns. For example, being held back makes girls more likely than boys to drop out of school at an earlier age (AAUW 1992, 49). Female dropouts are less likely to return to school and have much higher poverty rates than male dropouts, reflecting their lower earning power relative to males with the same educational attainment (AAUW 1992, 48). Although more girls than boys take college achievement tests, boys outperform girls on most of the tests, and win most of the scholarships tied to high performance on these tests (Sadker 1996). SAT tests themselves are biased against girls, in that they underpredict girls’ college performance, and overpredict boys’ performance (AAUW 1992, 56). Although more women attend college than men, they are concentrated in low-prestige, sex-segregated fields such as education, nursing, and library science that lead to lower-paying jobs than what college educated men--and even often men with only a high school degree--can expect. Men dominate majors, such as engineering and computer science, that lead to high-paying careers (AAUW 1992, 27, Sadker 1996). When girls suffer more from the same bad educational outcomes, and benefit less from the same good educational outcomes, this is a legitimate cause for concern, even if girls suffer the bad outcome less frequently than boys, and enjoy the good outcome more frequently.

Falsely claiming that feminist researchers care only about girls. Sommers calls feminist researchers “girl partisans” who have failed to advocate help for boys’ educational problems, at least until very recently (122, 125). This is false. Myra and David Sadker have been designing educational reforms on behalf of boys since the 1970s (Sadker 1996). The 1992 AAUW report argues that schools need to improve education for boys in numerous ways. It warns that boys may be misplaced and stigmatized in special education programs because their behavioral problems are misdiagnosed as learning disabilities (AAUW 1992, 19-20). It suggests that sexist cultural norms may depress boys’ reading scores by discouraging them from reading fiction (AAUW 1992, 23). It urges schools to encourage boys in this area, and in other areas where boys avoid acquiring the skills needed for success because these skills are coded “feminine” (AAUW 1992, 30, 85). It observes that teenage fathers have been neglected by schools, although they have poorer educational outcomes than other teen boys, and advocates educational programs to help them complete school and acquire the parenting and work skills they need to support their families (AAUW 1992, 40-1). It complains that teachers interact less frequently with black boys, perceive them as less able, and reward them more for good behavior than academic achievement (AAUW 1992, 70-1). It highlights the sexual harassment of boys as well as girls, and urges schools to take active steps against it (AAUW 1992, 73-4). It observes that vocational education programs have had “little success in increasing the earnings of men” and advocates reform to better serve boys and girls alike (AAUW 1992, 42-4).

Special pleading. In the few places where Sommers acknowledges gender gaps disfavoring girls, she tries to explain them away by appealing to assumptions that do not bear up under scrutiny. For example, she excuses teachers’ sexist double standard in the treatment of callouts by claiming that teachers are right to encourage boys’ callouts because “boys are less attentive” (107). This ignores the vast within-gender differences in tendencies to call out (Sadker, et al. 1991, 298). The boys who are not paying attention to teachers’ questions can hardly be the ones calling out answers to them. It is hard to see how they are encouraged by letting a few highly attentive boys dominate classroom discussion. Similarly, Sommers tries to explain away the fact that boys score more highly on both the math and verbal sections of the SAT by pointing out that the pool of girls taking the SAT reaches further down into the ranks of the disadvantaged (115). But boys dominate the ranks of those scoring most highly. A deeper pool of girl test-takers can explain a lower median score for girls, but not lower representation of girls among the top-scorers. Sommers explains this by claiming that there just are more male geniuses (116). This of course begs the question against feminist education researchers. Rather than addressing this obvious objection, Sommers tries to turn the tables on feminist education researchers by claiming that they are inconsistent. Observing that boys also dominate the low end of the ability distribution, she complains that if feminists see unfairness in the gender gap disfavoring girls at the high end, they should also see unfairness in the gender gap disfavoring boys at the low end (116). Quite right, which is why the AAUW did complain that boys may be overplaced in special education programs that depress their academic achievement (AAUW 1992, 19). Contrary to Sommers’ charge that the AAUW is composed of a bunch of “girl partisans,” the AAUW’s own position (1998, 1) is consistent and unequivocal: “when equity is the goal, all gaps in performance warrant attention, regardless of whether they disadvantage boys or girls.”

The editors of SFE presumably included Sommers’ paper because they took it to demonstrate that feminist research is dishonest, biased, and geared to reach “politically correct” conclusions. Sommers’ paper demonstrates none of these claims. Rather, it exemplifies the very flaws it claims to find in feminist research. Beyond that, its tone is hostile and contemptuous. Sommers is quick to infer insidious motives behind innocent phenomena. Her work has been subject to numerous searching critiques, sometimes occupying an entire issue of a journal (Sadker 1996; Auerbach et al 1994). The editors of SFE seem to be unaware of this. The also don’t seem to regard a harshly partisan tone and malicious accusations as a warning sign that the work in question may be unreliable. Perhaps this is because they fell victim to their own political correctness: assuming that, because a paper reached conclusions they found congenial to their own political agenda, it must be right.


American Association of University Women Educational Foundation. 1998. Gender Gaps: Where Schools Still Fail Our Children. American Institutes of Research. Washington, D.C.: AAUW Educational Foundation.

American Association of University Women. 1992. The AAUW Report: How Schools Shortchange Girls. Wellesley College Center for Research on Women. Washington, D.C.: AAUW Education Foundation and National Education Association.

Auerbach et al, Nina. 1994. “Symposium on ‘Who Stole Feminism?’ by Christina Sommers.” Democratic Culture 3(2).

Jaggar, Alison. 2000. “Ethics Naturalized: Feminism’s Contribution to Moral Epistemology.” Metaphilosophy 31(5): 452–68.

Kleinfeld, Judith. 1998. The Myth That Schools Shortchange Girls: Social Science in the Service of Deception. Washington, D.C.: Women’s Freedom Network.

Lee, Valerie. 1996. “Letter to the Editor.” Education Week on the Web, 7 August.

Lee, Valerie, Xianglei Chen, and Becky Smerdon. 1996. The Influence of School Climate on Gender Differences in the Achievement and Engagement of Young Adolescents. Washington, D.C.: American Association of University Women Education Foundation.

Pinnick, Cassandra, Noretta Koertge, and Robert Almeder, eds. 2003. Scrutinizing Feminist Epistemology: An Examination of Gender in Science. New Brunskick, N.J.: Rutgers University Press.

Sadker, David. 1996. “Where the Girls Are.” Education Week on the Web, 4 September.

Sadker, Myra, David Sadker, and Susan Klein. 1991. “The Issue of Gender in Elementary and Secondary Education.” Review of Research in Education 17:269–334.

Sadker, Myra, and David Sadker. 1984. Promoting Effectiveness in Classroom Instruction. Educational Research and Information Clearinghouse. Washington, D.C.: National Institute of Education.


[1]: I set aside consideration of Pipher’s and Gilligan’s research. Pipher is a clinical psychologist writing for a popular audience, not a research scientist. It is not proper to regard works of popular psychology as representative of academic research conducted by feminists. Gilligan is an academic, but her work has hardly been given a free pass within feminist circles just because she is a feminist. In fact, her work is highly controversial in these circles. I join the doubters among feminists for the same reasons tersely summarized by Jaggar (2000, 459).

[2]: Sommers makes the same error in accepting at face value the student perceptions reported in a 1997 MetLife study that found that ”more boys than girls. . .feel teachers do not listen to what they have to say” (119). She claims that the MetLife evidence “contradicted most of the pet ‘findings’ of the AAUW and the Sadkers.” But subjective perceptions must be tested against objective measures, which were the basis of the AAUW and Sadker research. The objective measures show that subjective perceptions in these matters cannot be trusted. Dozens of studies have confirmed the gender bias favoring boys in classroom interaction. In contrast with the 1992 AAUW report, which carefully cites studies that qualify or call into question various findings of bias against girls, Sommers trumpets studies that supposedly “contradict” the AAUW findings as definitive, without considering the weight of other evidence.

[3]: It should be noted that, in a changing world, it is improper to use data from the mid-1990s to “refute” claims about gender gaps made years before on the basis of data from the 1980s. The AAUW (1998) has forthrightly acknowledged that girls have made progress in reducing many gender gaps reported in 1992, thanks in part to the gender equity movement it helped to advance.