A few years ago, I worked with an assistant head who was phenomenally talented. During your time as a teacher, there will be that one colleague who just inspires you. For me, he was that person.

But this teacher’s methods were not always orthodox. One of the things he did was to divide his history classes by gender so that boys were taught separately to girls. I was dubious about this to begin with. But, after observing how he was able to tailor his teaching to each of the classes, I came to understand the rationale behind the decision.

When I asked him about equality, he simply told me that by splitting the classes he was able to achieve more equality. It seems counterintuitive, but, in a way, his split-gender class did create more equality in terms of outcomes.

International researchers, in particular, seem to think that there is a strong case for dividing genders in the classroom. Dustmann et al (2017) support the idea that split classrooms are more effective than co-ed classrooms. And Michael Gurian et al, in their book Successful Single-Sex Classrooms (2009), promote the idea that teaching boys and girls separately is an effective way to close a gender gap.

Both groups also allude to the notion that, by tailoring our approaches accordingly, teachers are able to create more equality. But this isn’t common practice in mixed-sex schools in England - and such approaches are often met with harsh criticism.

Split-gender classes weren’t something I thought about again until recently, when, as a lead teacher of English, I sat with the head of department and looked at our current and projected departmental data, alongside our data from last year. We quickly realised that there was a glaring issue: shortcomings in the achievement of the boys in our cohort. Their results weren’t hugely out of line with the national average, but they were notably lower than the girls’ results.

This raised the obvious question: what could we do to raise boys’ attainment and sustain those gains until the summer exams?

Grasping a thorny issue

Historically, girls do tend to outperform boys academically, but closing the gap in this particular cohort was a priority. The boys were simply not achieving to their potential and we needed to intervene.

As a department, we racked our brains to come up with a way to raise the attainment of the boys and make up for the deficit in our data. A lot of ideas were bounced around about intervention and extracurricular sessions, but we felt that these approaches wouldn’t be feasible in the long term. The complexity of re-weighting classes to create a higher boy-to-girl ratio in the top sets was another reason to rule out this option.

It was only when we went back to the drawing board that someone made the suggestion of splitting classes by gender and teaching them separately.

The simplicity of the idea drew a lot of scepticism. Ethically, of course, the concept of dividing by gender screams draconian archaism; merely considering it was enough to spark concern about the criticism it might attract from parents, students and the wider educational community.

There was also the fact that gender is fluid - some pupils identify differently to the sex they were assigned at birth.

Research conducted by the former Department for Children, Schools and Families (2009) rails against the idea of splitting classes by gender or varying teaching for either gender, so making the decision felt like a risk. But it also felt right.

Like any good researchers, we considered a variety of approaches. The decision to teach split-gender classes for English GCSE was a big statement and we wanted to ensure that we were able to achieve our goal of raising boys’ attainment while continuing to drive the girls forward, too.

So how did we do it? Logistically, it wasn’t especially difficult; besides the rerooming and timetabling, the practicalities weren’t a challenge. Winning the hearts and minds of the students, however, was more of an issue.

Like most young people, our students don’t like change. None of them had ever been in a single-gender class and we found ourselves facing a lot of questions before the shift took place. To overcome these issues, we were transparent with students. We told them that there were shortcomings overall in the data; we showed them the research suggesting that split-gender classes were beneficial; and we made it clear that we, as their teachers, thought this would be the best approach for them. In addition, they became participants in our own small-scale study.

As a team, we had done some reading on approaches to teaching different genders and then compiled our research. We agreed that the content and curriculum would remain consistent and only the teaching approach would vary. For ethical reasons, we were determined to teach the classes exactly the same content (depending on ability, of course), but we wanted to tailor the approaches to best suit each of the genders. This was what we hypothesised would be the lynchpin of our success.

Starting with the 2020 Vision report (Teaching and Learning in 2020 Review Group, 2007), we noted the importance of personalising learning for genders. We looked closely at the approaches that were found to benefit boys, and used them as a starting point for our research. They weren’t hugely groundbreaking and, from the outset, the simplicity and evidence contradicting them meant that we were sceptical. But we reminded ourselves not to discount anything until we had applied the ideas in our context.

The report noted that boys tended to place lower value on language and literacy than girls, so we decided to plan phases into the lessons that built on vocabulary and allowed for opportunities to apply new terms and ideas. It’s no surprise that boys were found to benefit from highly active, multisensory work. So, again, we made sure the learning involved elements of this to engage the boys fully. But it didn’t feel like these approaches alone were going to solve our issue.

Digging a bit deeper, we discovered a lot of research had been done about the thinking around these types of personalised gender approaches. Cleveland (2011) conducted a qualitative survey of the characteristics and approaches that boys responded best to, asking them a series of questions about school and how they felt about certain aspects of their learning. Having reflected on Cleveland’s research, we conducted a survey using similar (but not identical) questions.

• What do you struggle with in lessons?

• What do you think helps you succeed in lessons?

• What do you love about school?

• What do you dislike about school?

• What would you say you are successful at?

• Which aspects of learning do you find most difficult?

The findings allowed us to tailor our approach - something that, as a department, we felt more ethically and morally comfortable with than using strategies such as those put forward by the Teaching and Learning in 2020 Review Group, which felt too stigmatised and stereotypical, especially considering that they were proposed so long ago and were so highly contested. Our findings were as follows.

Boys struggled most with vocabulary and writing. They felt that building up levels of challenge and success allowed them to succeed most overall, and the thing they loved about school was winning at new things. Interestingly, the boys in particular were most concerned about failing - something that they noted made the idea of winning (or not winning) threatening. The boys felt more successful at English language because it was logical; they said they felt that expressing their feelings about texts and extracts was the most difficult thing about the subject.

Girls struggled most with having to stick to rigid structures when answering questions. We found that they enjoyed challenge from the outset of a lesson and that they loved being seen to succeed. The girls’ biggest concern was not knowing how good they were at something until an assessment or formally marked piece of work confirmed it. The girls were most proud of their previous performances and they felt like they used their feedback to improve their learning. The main difficulty in English for the girls was remembering everything.

From this qualitative basis, we were able to draw five key elements that to use to differentiate the learning for each of the gender groups.

Boys

• Provide sequential lessons with a growing degree of difficulty.

• Embed literacy reminders and tasks.

• Offer additional opportunities to interleave creative writing.

• Use low-stakes tests to build recall.

• Create logical systems for answering literature questions.

Girls

• Start the lesson with a high-level activity, including elements of recall.

• Heighten live feedback opportunities in lessons for longer tasks.

• Frequently celebrate and revisit learning from different units.

• Provide models to support answers, but ensure freedom in their own work.

• Always use mark schemes to allow banding/levels (not grades) with work.

Besides these agreed approaches, we endeavoured not to vary the topics or the overall outcomes of the lessons to ensure consistency. We also taught the lessons as we normally would, using the same fair behavioural systems.

Gap is narrowing

Does this approach really work? It is still early days in our trial, but there has already been a notable effect on attainment. Across the board, on average, there has been an increase of half a grade in English literature and language for the boys. Subsequently, the gap has been slightly closed on the girls (who are also performing more positively than before).

What’s more, the data seemingly indicates that boys who are eligible for the pupil premium have responded especially well to the split-gender classes, showing that those in this sub-group have particularly benefited from the tailored teaching approaches.

There is still a deficit among boys with lower prior attainment, but we plan to refine our approach by going back to our initial findings and seeing whether we can further tailor their learning to suit them as a group.

In a follow-up wellbeing survey that we conducted recently, the students spoke incredibly positively about the way in which their English classes were split. They made a number of comments about feeling more comfortable and more willing to contribute in class discussions. Their inclusion in our research also meant that they felt valued during the process - we got buy-in.

With such good initial results, you might think that split-gender groups would be something that we should be applying across the board. But there are a lot of merits in mixed-gender classes and the research suggests that, for the most part, this is the best way to teach.

Whether we look into this approach with other groups in other subjects is still open for debate, though.

I am personally not completely convinced that split-gender groups are always best. But when it comes to exam preparation, splitting the genders does seem to be a great way to really focus students’ minds.

Adam Riches is a specialist leader of education and lead teacher in English

This article originally appeared in the 15 March 2019 issue under the headline “Split decision”