Girls Good, Boys Bad? Revisiting the Old Arguments About Single Sex Education

Whenever I hear people arguing that "progressive" social scientists really just find what their "liberal" biases lead them to find, I think about single-sex education. There has long been a sense, more of a "gut feeling" among some, that girls that exclude boys provide girls with a better educational experience for a wide range of reasons. Perhaps the most compelling idea is that with boys out of the classroom, girls will have a chance to shine that they don't otherwise have in our male-oriented and dominated society.

But evidence that single-sex schools are better for girls has been extremely difficult to find, even by those who really want to find it. I'm no expert on this, but the last time I looked, a few years ago, I believe that the general conclusion in the literature was that single sex schools didn't seem to lead to any significant difference in outcomes for girls. (I know much less about the literature around single-sex environments for boys.)

Slate.com now cites a recently published study (the complete study is here) giving evidence that boys are corrosive on classroom environments. The study looked at nearly half a million students in Israel's education system, with large numbers of observations.

As the authors note in their abstract:

Our results suggest that an increase in the proportion of girls leads to a significant improvement in students’ cognitive outcomes. The estimated effects are of similar magnitude for boys and girls. As important mechanisms, we find that a higher proportion of female peers lowers the level of classroom disruption and violence, improves inter-student and student-teacher relationships as well as students’ overall satisfaction in school, and lessens teachers’ fatigue. We find, however, no effect on individual behavior of boys or girls, which suggests that the positive peer effects of girls on classroom environment are due mostly to compositional change, namely due to having more girls in the classroom and not due to improved behavior of peers.

Their conclusions include the following:

An examination of the underlying mechanisms of the gender peer effects shows that a higher proportion of girls in the classroom lowers the level of classroom disruption and violence, and improves inter-student and teacher-student relationships as well as students’ satisfaction with school. It also significantly alters teaching methods and lessens teachers’ fatigue and feelings of burnout, . . . although it does not affect their overall work satisfaction. On the other hand, we find no evidence that having more girls in a class leads to clearer and more enforceable disciplinary rules at school.

Interestingly,

The estimates of the effect of the proportion of girls on student’s (self-reported) violent behavior, disciplinary problems, and study effort show no systematic or significant relationship, suggesting that much of the improvement in the classroom environment associated with a higher proportion of girls is due to a change in classroom gender composition and not to changes in individual student behavior.

I have a number of questions about this study, of course (which, it is important to note, is a "working paper" and not a peer-reviewed publication). For example, the focus on "girls" in one particular (collection) of cultural context(s) makes me wonder about how much of the effect, if it is accurate, results from a particular way of raising boys and girls (and of responding to them in the classroom) than from being "boys" in some absolute sense. What does it mean to be a "boy" absent a culture? What exactly does it "mean" that the driving "mechanism" is gender writ large as opposed to changes in individual student behavior? As usual with statistical studies like this, the findings are really not findings at all. They simply raise questions. What exactly leads to these results, to the extent that they are accurate? What are the interactive mechanisms on the ground that produce the responses on the surveys they analyze? Etc.

There are also wider issues about how one presents data like this, and what effect this paper may have on quite contentious discussions going on about the relative achievement levels of boys and girls in school (and their later success) across different groups.

In any case, if these conclusions hold up, I think they are significant and important for us to think about.

Let me conclude with the usual qualifications. Gender and schooling is not my area of focus, so I hope others more studied in this area will correct any failings. Also I have not attempted to wade through the extensive justifications and sub-arguments of this paper (and I'm not really equipped to do so with respect to the statistical aspects). Others may want to look more closely. As usual, I raise this for discussion.