AT Conyers School in Yarm, north-east England, we decided to explore gender setting. "We" are the 12 science teachers in an 11 to 18 comprehensive with 1,200 pupils. Although half the department staff are women, we still contend with the persisting belief (reflected in the balance of our A-level students) that science is a male thing. And we also wanted to address the problem of boys' underachievement.
Last summer we carefully divided the 208 Year 8 pupils into two lots, as equal as possible in terms of achievement, and set up our experimental groups for teaching in Year 9. One half of the year group was to be our control, continuing to be taught in mixed classes. These pupils were divided again into two broad bands giving two upper ability sets and two lower ability sets.
The other half of the year group was also divided into four sets - but this time upper-ability boys and lower-ability boys, upper-ability girls and lower-ability girls. As far as possible, each teacher involved was given a mixed and a single-sex group of comparable ability. For instance, if you had an upper-ability mixed group, you would also have either an upper-ability all-boys group or an upper-ability all-girls group. As Year 9 classes are taught by two teachers, as far as possible each group was assigned one male and one female teacher. Timetabling constraints meant that this was not always possible and the lower-ability girls' group and one lower-ability mixed group were taught by two women teachers.
We all learned some things about teaching boys and teaching girls which will be useful in our customary mixed classes. Boys most definitely needed short, structured tasks, with plenty of variety. Girls were much happier working on a more extended activity. We all found how much boys dictate the pace in mixed classes even if we try not to let them. Girls were seen to be much better at group tasks, being much more skilled in the co-operation, communication and negotiation.
The outcome was surprising. At around Christmas time the whole year group sat a common test, but the results were alarmingly inconclusive. Had we set a test which was inappropriate (though we couldn't say why) or had the experiment just not been running long enough?
But in June, the SATs results showed clearly that the people who did best were the girls taught in mixed sex classes. More detailed analysis, based on improvement on key stage 2 results, showed that at the upper end of the ability range the girls in mixed classes had indeed done better than either the girls in the all-girls top set or any of the boys. However, among the lower-ability girls and boys, teaching in separate-gender classes showed a more positive effect. Maybe gender setting works for some people some times. But can it always work? We will repeat this experiment for the next two years to see if we can find the answer. Meanwhile, we are reflecting on any gains in pupils' confidence and enjoyment.
Patricia Miller is head of physics amp; KS3 science at Conyers School, Yarm, Stockton-on-Tees