From the Middle School High School Principal

Why we do not stream maths, and why some 'intuitive' ideas turn out not to be true

If you want a good example of why intuitive ideas are not always true, ask a child under four where wind comes from.  The vast majority, never having heard it from anyone else, will say 'trees'.  

When I was first told this by educational researcher, Dylan Wiliam,  I thought 'nonsense!'.  I rushed home and asked Luca who was then four, 'Luca, where do you think wind comes from?'.  He looked confidently at a mango tree growing in Jalan Kelempong Park, Singapore and said 'Trees'.

The point obviously is not really about trees. The point is that often what our intuition tells us is common sense, turns out not to be.  

To most people streaming maths is common sense intuition.  Surely some children need hard problems and to go really fast, and others need less content and a slower pace.  As it is easier to do this if you stream classes (e.g. divide learners by perceived ability or 'intelligence'), then surely you should stream your classes?

The research by the UK's educational endowment fund (a major educational research funder aligned to the UK government, and generally in favour of more traditional approaches) found that having streamed classes is not 'neutral' but it reduces outcomes, particularly for 'low-ability' students.  Just read that again.  Streaming sets so that slower learners can go at a slower pace in theory 'more suited to them', actually means their learning is negatively affected.  That is slightly counter intuitive.  Some studies show that it is possible in some circumstances that streaming might have a near negligible effect for higher performing students in certain circumstances, but this is so insignificant, that the average effect is that streaming to match ability lowers achievement.  These studies were all specific to maths teaching, which is the area where  people most often feel inclined to stream by ability.

If you look at lower ability students in streamed sets, research found that they fall behind on average 1-2 months per year based on research by organisations like the Sutton Trust.  That is a stunning gap in their achievement compared to where they would be in a class that was not streamed.

There are no meta studies which show that streaming will raise average scores.

How can our common sense be so wrong? Some reasons might be...

• Streaming into 'low sets' can encourage a shared perception in both teachers and students that students should lower their sights in terms of academic achievement.  Both parties may 'live down' to the expectations.  
• Students in lower sets have no role models, and more able students miss out on the benefit from seeing how others make sense of problems, or even helping a peer to understand a concept which would consolidate their own understanding.  
• The range of skills, e.g. social and communication skills children might have, are not linked to maths skills - so having an artist who is not so good at maths but very good at visual communication, working on a project modelling mathematical rates of change with a child who loves maths and computing but finds communication hard, is a win-win.
• Similarly if you are grouping by perceived 'ability' you miss the opportunity to arrange classes in ways that are more sensitive to other kinds of diversity, social dynamics, needs etc.  So you overall get less well balanced and functional classes.
• Children being explicitly 'ranked and sorted' can undermine wellbeing in a wide variety of ways, and lead to fixed mindsets about ability.
• Children can feel a sense of 'conditional regard' or being judged in a prejudiced manner, rather than being given the opportunity to show what they are capable of.
• A 'top set' can give the feel that learning is a race, and remove the joy and curiousity of learning.  It can also produce a fixed mindset and fragile egos.
• If there is an alternative examination method, such as in our case the IGCSE core exam, which is designed as a modified exam - the danger is the teacher uses this as a modified curriculum.  That means the learner misses out on significant sections of learning they will need in future courses like the IB in grade 11 and 12 in our case.  That's why we teach the full curriculum you need to be ready for IB to everyone, but offer a modified 'core' or more challenging 'extended' exam as just one measure of what you have learned.  But we do not teach 'down' to an exam which does not represent the full range of the learning you actually need later.  By having a wider conceptual framework, it also makes it more likely you will do well on the modified exam.
• Children also flourish and develop at different rates, so even if you think you are streaming by ability, it is quite likely you are wrong, or turn out to be wrong, about who should be placed where.

So it turns out our intuition that streaming maths is a good idea, turns out not to be so true.  Which just leaves one question for all of us....

Where on earth does wind come from?

(This article was written after a conversation with the CEC rep chair Meta who asked for more information on why we don't stream in Grades 9 and 10 Maths and suggested sharing our thoughts.  I hope this helps with the rationale.  As mentioned above, both examinations continue to be an option, but not a curriculum.  The curriculum is a broad based and rich preparation for the IB, High School Diploma and indeed, life).