The evidence on ability grouping appears relatively well-known. The EEF Toolkit summarise the research findings thus:

Overall, setting or streaming appears to benefit higher attaining pupils and be detrimental to the learning of mid-range and lower attaining learners. On average, it does not appear to be an effective strategy for raising the attainment of disadvantaged pupils, who are more likely to be assigned to lower groups.

It appears that children who are deemed to be ‘low ability’ fall behind pupils with equivalent prior attainment at the rate of 1-2 months per year when placed in ability groups. Conversely, high attainers make, on average, an additional 1-2 months progress per year when they are set. There’s much speculation about why this might be. Here are a few popular theories:

  1. Low ability groups are assigned less capable teachers. Top sets are often seen as a reward, bottom sets a punishment. If low attainers are viewed as unlikely to make good progress then it might not make strategic sense to assign them your best teachers.
  2. When children are corralled together by ability, they learn that they are either ‘bright’ or ‘thick’ and then rise or sink to meet these expectations.
  3. Behaviour in ‘bottom sets’ prevents students from learning. I wrote about this here; it’s scandalous that some schools continue to allow bottom sets to be sinks of low expectations and poor behaviour.

Another more interesting idea is that setting may actually cause differences in ability. The late Graham Nuthall put it like this:

Ability appears to be the consequence in differences of what children learn from their classroom experience.

Even if not strictly true, this hypothesis is the one most likely to lead to equitable experiences for all children. There’s no doubt that some children are more intelligent than others, like most traits, intelligence distributes normally across a population. But, that doesn’t mean schools are especially good at identifying who is more or less able.

The biggest and most important individual difference between children is the quality and quantity of what they know. Let’s imagine a scenario where two students – Katie and Liam – join school mid year and need to be placed into sets. Katie has experienced successful phonics teaching and mastered decoding in Year 1, moving quickly to more interesting and sophisticated reading material. Liam on the other hand suffered with undiagnosed glue ear and was unable to properly make out the fine distinctions between different vowel and consonant sounds. Although he can decode, his ability is halting and laborious. Too much of his fragile working memory capacity is spent on sounding out letters with little left to spare for much in the way of higher level comprehension. Both pupils are assessed using a reading comprehension test; Katie scores well while Lim does poorly. As a result Katie is placed in the ‘top set’ and Liam in the ‘bottom set’. On the face of it this appears entirely reasonable, it could be the case that Liam is actually more intelligent than Katie but just knows less.

This might sound far-fetched, but Dylan Wiliam estimates that when tests are used to select children for ability groups “only half the students are placed where they ‘should’ be.” In other words, it comes down to a coin toss as to whether students are placed in the right class for their ability.

The EEF report advises schools to “recognise that a measure of current attainment, such as a test, is not the same as a measure of potential.” This is, of course, true, but more important, what happens to children’s potential once they have been set?

Let’s return to our fictitious students. In the top set, Katie is given more challenging material at a faster pace. Her early advantage is compounded. Liam though is given much more simple things to do at a slower pace ensuring that, relatively speaking, he knows less and less. The more we know, the more we can think about and the cleverer we become. Children’s experience in school determines, to a large extent, their ability. After all, no one rises to a low expectation.

Of course, none of this is fate; the research reports what has been not what could be. Conceivably, a school my have designed an approach to setting in which middle and low attainers are not held back, but we can be reasonably sure what is likely to have to children if a school’s approach to setting is broadly similar to those that have gone before.

My advice is that for the most part, we should delay grouping pupils by ability for as long as possible. If some children are holding back the progress others of others because they have not mastered basic, foundational knowledge, they should be taken out for short durations to ensure they taught what they need to know and then returned to normal lessons. Of course there are some children within the system whose needs are so acute that they need special provision; none of these comment should be seen as applying to them.