A new study of 87,921 students [KR]

A new study of 87,921 students: students RANDOMLY assigned to single-gender schools had better outcomes than students assigned to coed schools.

 

Leonard Sax (http://www.singlesexschools.org/)

Researchers at the University of Pennsylvania traveled to Seoul South Korea, because in Seoul, students are RANDOMLY assigned either to single-gender or to coed high schools. The assignment is truly random, and compulsory. Students cannot “opt out” of either the single-gender format or the coed format. This policy of random assignment was instituted in 1974 specifically to prevent clustering of students from particular backgrounds at particular schools.  In recent decades, many Korean school districts have loosened the policy and they now allow parents to express preferences or to “opt out” of particular schools. But not in Seoul. In Seoul, it’s still a true random assignment with no opt-out.

          The scholars from Penn recognized that the random nature of the assignment creates the opportunity to compare single-gender schools with coed schools, without the usual confounding variables which would accompany any attempt at a similar comparison among North American schools. All the schools in the study are publicly-funded; none of them charges fees or tuition. The researchers found no differences between the single-gender and the coed schools in terms of teacher quality or in teacher training. Class sizes in the boys’ schools were no different than in the typical coed school, and class sizes were actually slightly larger in girls’ schools than in the typical coed school. There were no differences in socioeconomic background or prior academic achievement between students attending single-gender schools and those attending coed schools.

          What were the results? Girls attending girls’ schools were significantly more likely to attend a 4-year college compared with girls attending coed schools (Cohen’s d = 0.5, p < 0.01). Likewise, boys who graduated from boys’ schools were significantly more likely to attend a 4-year college compared with boys who graduated from coed schools (Cohen’s d = 0.8, p < 0.01). All these effects remain significant after controlling for eligibility for free school lunches, prior academic achievement, and other demographic and student parameters. Boys at boys’ schools also earned significantly higher test scores compared with boys at coed schools; likewise, girls at girls’ schools also earned significantly higher test scores compared with girls at coed schools. The authors conclude:

 

Our analyses show that single-sex schools are causally linked with both college entrance exam scores and college-attendance rates for both boys and girls. Attending all-boys schools or all-girls schools, rather than attending coeducational schools, is significantly associated with higher average scores on Korean and English test scores. Compared with coeducational schools, single-sex schools have a higher percentage of graduates who moved on to four-year colleges.

 

I am grateful to Dr. Christina Hoff Sommers for alerting me to this article, which will be published later this year in the journal Demography. You can read the final manuscript as it will be published later this year at the publisher’s web site, http://link.springer.com/article/10.1007/s13524-012-0157-1. However, the publisher will charge you $40 for a PDF. Fortunately, the authors have posted the full text of their article, in manuscript form, on the University of Pennsylvania web site, at no charge: go to  http://repository.upenn.edu/psc_working_papers/15/, then click on “Download” on the right. I have compared the two versions and I cannot detect any differences.

 

This article is important. As the authors observe, this study is the first large-scale study of students RANDOMLY assigned to single-gender and coed schools. My only concern with the article is with its underlying premise: namely, that either single-gender or coed must be “best.” As I have mentioned in these updates previously, I think that premise is fundamentally mistaken. In my visits to more than 300 schools over the past 12 years, I have found that the single-gender format is better for some students, and coed is better for others. For me, the most interesting question in this field is: which student is most likely to benefit from the single-gender format – and which student is NOT likely to benefit?

          I know of nobody who has any interest in pursuing this question, which to me seems to be the key to any further progress in this field. Most leaders of single-gender schools have very little interest in this question: most of them (though not all) would not welcome any research which would identify students who should NOT apply to their schools. And the zealots at www.coedschools.org, led by Professor Diane Halpern, Professor Janet Hyde, Professor Rebecca Bigler & Co. will never concede that ANY student would benefit from ANY single-gender format.

          You might be amused to know that in our ‘debate’ in Houston this past October, I asked Professor Bigler – distinguished professor of psychology at the University of Texas at Austin – whether she is opposed to the Girl Scouts. She answered, “Yes.” She has no objection to the idea of Scouts in general, but she sees no justification for ‘segregating’ girls from boys in the Scouts. All kids, girls and boys, should go camping together, have all their activities together, at every age, no exceptions. Even addressing children as “girls and boys” or “ladies and gentlemen” is wrong. That’s Professor Bigler’s position.

          I had nothing to say to that.


Leonard Sax

 

LINK-1

LINK-2