Stereotype threat, risk-taking and healthy competition

• Andrew Hill’s research has found that in American students aged under 16, grades in mathematics and science are negatively impacted by the effect of opposite gender friends. He suggests that girls may benefit from single-sex classes in traditionally male-dominated subjects such as mathematics and science, stating that the negative effect of co-educational classes for girls “is consistent with gender socialization effects in the existing developmental psychology and economics of education literature” showing that “adolescent females may shy away from competition and perform less well in mathematics in the presence of males” (Hill, 2015, p. 168).

• Booth and Nolen (2014) summarised the evidence gathered from two “real-stakes gamble” experiments investigating the risk-taking preferences of males and females in single-sex and mixed environments at school and university. They concluded that girls from single-sex schools “are as likely to choose the real-stakes gamble as boys from either coed or single-sex schools, and more likely than coed girls” (section 3.2.3) and that in a co-educational university environment, girls randomly assigned to single-sex classes were “significantly more likely to choose the lottery than their counterparts in coeducational groups, and the magnitude of the effect was quite large” (section 4.4). Booth and Nolen proposed that there may be a reduction in stereotype threats in single-sex environments where females do not feel “inhibited by culturally-driven norms about the appropriate model of female behaviour” (that is, avoiding risk), or that in all-girl environments, friendships formed with other females may enhance their confidence, “leading these women to feel more comfortable in making risky choices than women in coed classes” (section 6).

• Booth, Cardona-Sosa & Nolen (2014, pp. 126) conducted a risk-taking experiment in a co-educational university in which males and females were randomly assigned to single-sex or co-educational groups, finding that “after eight weeks in a single-sex environment, women were significantly more likely to choose the lottery than their counterparts in coeducational groups” and that these results are robust for IQ, personality type and a number of sensitivity tests. “No such result was found for men in the single-sex groups” (p. 140).

• Booth, Cardona-Sosa and Nolen (2013, p. 16) discussed ‘stereotype threat’, when girls are stereotyped as “bad” at something (such as maths or economics), and conjectured that girls would do better in all-female classes where there is a reduction in “psychological threats caused by studying with males”. They concluded that the results of their single-sex class experiment at Essex University, which showed that girls assigned to single-sex classes outperformed those in co-ed classes, demonstrated that “there is evidence that all-female classes have a direct effect on pass rates and course scores as predicted by the reduction of stereotype threat” (p. 18).

• Booth and Nolen, writing about competitive and risk-taking behaviour in girls, stated that: “…the bulk of our evidence suggests that a girl’s environment plays an important role in explaining why she chooses not to compete. We have looked at the choices made by girls from single-sex and co-ed schools and found that there are robust differences in their behaviour: girls from single-sex schools behave more competitively than do coeducational girls” (2009b, p. 20; also see, 2009a).

• Summarising recent research on the effect of gender on competitiveness, Buser, Niederle and Oosterbeek (2014, p. 1411) state: “Experiments have shown that for women both the performance in as well as willingness to enter competitive environments is reduced when the competition group includes males (Gneezy, Niederle, and Rustichini 2003; Balafoutas and Sutter 2012; Niederle, Segal, and Vesterlund 2013). Similarly, Huguet and Regner (2007) show that girls underperform in mixed-sex groups (but not in all female groups) in a test they were led to believe measures mathematical ability.”

• Suzanne Link’s 2012 study of the 1999 Trends in Mathematics and Science Study (TIMSS) data for South Korean middle schools found “positive effects of single-sex schooling for girls” in mathematics (p. 2). She writes that the effects in mathematics “are not only highly statistically significant and non-negligible in their magnitude, but also highly relevant since math[s] performance is consistently linked to future earnings” (p. 2). Link found that the test score gender gap is “especially large for girls from low parental support backgrounds who attend coeducational schools”, suggesting that “these girls might be somehow harmed by the presence of boys when learning in a stereotypically male subject such as math[s]” (p. 17).

• A 2012 study by Doo Hwan Kim and Helen Law of the gender gap in mathematics found that in South Korea, “single-sex education shields girls from exposure to the prevailing gendered society and provides a favourable atmosphere for their success in maths and science”. In addition, they wrote, their analysis of Korea’s 2010 College Scholastic Ability Test (CSAT) results revealed “the advantage of attending a single-sex school”, stating that the “mean scores for both girls’ and boys’ schools were higher than those for their co-ed counterparts in all three major CSAT subjects: Korean, English and maths”.

• A 2011 study by Titze, Jansen and Heil compared accuracy in a test of mental rotation in girls from single-sex and coeducational schools. In previous research, this test has consistently been performed better by males than females. Titze et al.’s study found that girls at single-sex schools outperformed girls at co-educational schools in the mental rotation test. These results add to the body of evidence that single-sex educated girls are less “gender-stereotyped” in their development, than girls at coeducational schools (p. 5).

• Cherney and Campbell’s 2011 paper found that “girls in single-sex schools had significantly higher self-esteem and higher achievement motive” than girls in co-educational schools (p. 721). They also found – even after controlling for prior grades, age, maths anxiety and domain identification – that both boys and girls from single-sex schools “performed significantly better” on the maths test than their co-educational counterparts (p. 721). Girls at single-sex schools achieved mean test scores of 5.1 in maths compared with 4.2 for girls from co-educational schools (p. 719). Interestingly, girls showed “superior” maths performance under stereotype threat conditions (when their test instructions stated that boys typically did better on maths tests than girls) compared with girls who read the “no threat” instructions (p. 721). Cherney and Campbell discuss factors that may account for this, including that girls in single-sex schools have “predominantly female math and science teachers” and that “it is possible that girls in single-sex schools are more likely to perceive their math ability as malleable” than girls in co-educational schools (p. 722).

• In a report authored by Dr Richard Holmgren for the US National Coalition of Girls’ Schools (NCGS) analysing the 2013 High School Survey of Student Engagement (HSSSE) conducted by Indiana University — which surveyed 2013 girls attending NCGS member schools, 5210 girls from co-ed independent schools and 5741 girls from co-ed public schools — it was found that 95.5% of girls at NCGS schools agreed or strongly agreed that they actively participated in classroom discussions. Of these, 41.5% of girls at NCGS schools strongly agreed that they were engaged by teachers in classroom discussions, in contrast to 33.3% of girls in co-ed independent schools and 12.9% of girls in co-ed public schools (Holmgren, 2014, p. 4).

• In 2011, Minjeong Lyu and Diane Gill’s study of co-educational and single-sex physical education classes in in South Korea concluded that “same-sex classes lead to higher competence, confidence, achievement, enjoyment and effort” and that “females are more affected by class type than males” (p. 256). They write that, in spite of the argument that co-education gives students “the opportunity to understand each other and reduce inaccurate gender role perceptions, there are negative side effects in coeducational physical education” (p. 256). In sum, “same-sex classes may be the better teaching environment in adolescent physical education, particularly for girls” (p. 257).