Why can’t boys be…Well, more like girls?

In Uncategorized by Alex Quigley3 Comments

[This blog first appeared on the Huntington Research School website. Take a look HERE and sign up for our monthly newsletter HERE, so that you get access to the brilliant blogs of our entire team.]


Schools are complex places full of blood, sweat and break-times. Making sense of all of the complex factors that determine the success of our students is a devilishly difficult task, like the proverbial needle-seeking in a haystack.

When faced with such difficulty, whilst being tired and time-poor, we instinctively look for short-cuts to help us. One such shortcut to action is data. Schools are awash with data.

Despite all of the data, seldom do school leaders and teachers gain deep, sustained training about using it well. In the busy hubbub of the school week, we then become prone to all-too-human biases. Our short-cuts become deeply rooted in our daily experience. One common short-cut is stereotyping. Our biggest, more prevalent stereotype? Well, it has to be gender doesn’t it?


Educational iconoclast, David Didau, has written a brilliantly provocative blog on how we use gender as a shortcut to action in schools, whilst drawing into question how prevalent it is in our thinking and school organisation:

“I’m not suggesting gender has nothing to do with attainment – it probably does have some bearing – but maybe a lot less than we’re inclined to believe. And the extent to which gender might be causal is more likely due to cultural rather than biological causes, as this article makes clear. Our best bet is probably to insist on high expectations for all students and not let boys get away with being ‘just boys’.” David Didau, ‘What causes the gender gap in education?

What is particularly helpful in Didau’s blog is his example of Mr Garvery, who, when presented with data on GCSE performance, noted the gender differences, but this led to him failing to see more significant differences, such as teacher and student attendance.

The Mr Garvery example is a helpful example of ‘salience‘, another of our instinctive mental baises. It is rather like the reality of shark attacks. Thought they loom large in our imagination, the likelihood of our dying by shark attack is infinitesimally small compared to death by drowning, and yet our psyche is scarred with the dangers of sharks!

great white shark

“It was the boy shark that done it, Miss”

Gender differences in our society are clearly so pronounced and salient, from unequal pay, to many more differences and inequalities (predominantly tipped against women), that it is natural to seek out gender as a cause for differences in the classroom, but we need to be better at scrutinising and analysing data and being aware of our intuitive biases.



Recently, analysis of our school data showed that ‘we had a gender problem‘. The boys weren’t working as hard, and their behaviour comments stacked up against their female counterparts. All very familiar. What could we do? Should we have a special ‘male motivation’ group? Was the curriculum or assessment not ‘boy friendly’? Perhaps, we needed to have an assembly telling their boys to buck their ideas up? What if mixed gender groupings were holding girls back?

When you digged a little more beneath the salient and stereotypicalheadlines, the gender issue wasn’t really the issue at all. We had some subject specific issues in terms of under-attainment and behaviour, but those were not long-term trends. Like Mr Garvery, we were finding that our assumptions about large swathes of the boys in our GCSE groups simply weren’t true.

Typically, we analyse the data of year groups, starting with the usual Pupil Premium and disadvantage indicators. Then we quickly shift to sub-groups like gender, then FSM, SEND, EAL and many more. Crucially, we spot those salient patterns – remember the shark – such as gender. We should be very wary of sub-group analysis. With small cohorts (pretty much every school has a small cohort when it comes to the study of statistics), then sub-divided into even smaller cohorts, we create false results due to an “insensitivity to sample size”. Put simply, the smaller the sample, the greater the variation.

Where do we start then? Well, we know that gender can correlate with student outcomes, but we need to analyse an array of student and school level factors that are likely to prove more important. If you read the excellent Edudatalab website, you can crib up on how school with the most able prior attaining students still make most progress; the influence of ethnicity, EAL and long-term disadvantage; trends related to birth order, and much more.

Dr Becky Allen, data expert and formerly leader of Edudatalab, has sage advice for every teacher and school leader:

“Beware of looking at gaps in attainment between two groups in your school. You will almost always find some difference in attainment between one group and another purely due to chance. Beyond this, it is far more likely that your gap arises for reasons that are entirely outside the control of your school, such as home circumstances.

If you are still want to do something to help your boys, then the current research evidence says the most important thing to pay attention to is school behaviour, for boys learning appears to be significantly affected by classroom disruption. Of course, in lowering classroom disruption, you might find your girls benefit too and so your ‘gap’ fails to close much! After all, there is more that is similar in the minds and personalities of girls and boys than there is difference.”

When we are reflecting upon gender patterns in our schools, we can ask to following questions to ensure that we don’t fall for the salient stereotype that ‘girls are better than boys’, or that all boys need a “boy friendly curriculum”, or other simplistic stereotypes drawn from our analysis of data:

  • When reflecting on gender patterns in data analysis, what sub-group analysis of other characteristics can you explore alongside gender?
  • Are sub-groups EAL, age, disadvantage, prior attainment, absence rates etc adding useful indicators to our data analysis? Careful though, remember: 
  • How is the size of the data-set skewing the sub-group analysis and creating false headlines?
  • How reliable and valid is the data I am analysing?
  • Are there three year trends I can explore so that I am not swayed by the latest exam data or similar? 


(Thank you to Dr Becky Allen – and her Edudatalab team – from pushing forward the profession to better analyse data and see beyond the salient headlines)


  1. Pingback: Girls, Girls, Girls – Peer Effects in the Classroom

  2. Pingback: Alphabetical Signposts to Teacher Excellence – B – Teach innovate reflect

  3. Pingback: What Teachers Tapped This Week #26 - 26th March 2018 - Teacher Tapp

Leave a Reply