It’s probably fair to say that most discussion of concepts like setting, streaming and banding in a-level sociology focus on things like the basic principles involved or the social and psychological consequences of different kinds of “ability grouping”. While this is, of course, a perfectly valid set of concerns (pun sort-of intended), there tends to be little or no discussion of the actual mechanics of the process.

While the benefits and drawbacks of something like “ability setting” are endlessly debated, the assumption on both sides has tended to be that this is something of a technical process of allocation: students are measured at Key Stage 2 (when they are 11 years old and in the last year of primary education before moving to secondary school) and on the basis of their KS2 results are allocated to different sets. These are usually of the “High”, “Middle” and “Lower” variety.

Again, while there has been a measure of debate about the validity of both the Key Stage tests themselves and the testing regime, quantifiable subjects – Mathematics in particular – have largely escaped criticism precisely because, whatever the debate about the merits or otherwise of setting, at least it’s based, in this subject at least, on measurable, quantifiable, criteria.

A recent study by Francis et al (2019) has, however, cast a degree of doubt on both the allocation process and, by extension, the claim that setting into secondary subjects is purely based on academic achievement.

Their argument is that the process is actually infused with a range of gender and ethnic biases.


If you just want the basics:

The study looked at data on nearly 10,000 (9,301) children drawn from Year 7 (12 – 13 years old) students currently in 46 randomly selected secondary schools.

More nuanced:

The schools were a subset of the 127 schools that had originally signed-up to a UCL “Best Practice in Setting” trial that ended (inconclusively) in 2018.

The 46 schools had originally been randomly assigned to the “business as usual’ control group in the trial. This meant they had simply carried-on with their normal setting and teaching practices without making any of the changes and adjustments required of the non-control schools.

Although schools were originally recruited into the trial through “a mixture of volunteer and direct ‘cold call’ approach sampling” before being randomly assigned to control and intervention (i.e. those who were to be instructed in “best practice”) groups, it’s difficult to know how, if at all, this potentially biased the sample in the setting study.

On the one hand, it may be the case that those schools who agreed to take part in the Best Practice study were confident they were already broadly following “best practice”.

On the other it may be the case these were schools who were not confident they were following best practice and were seeking guidance through their participation.

Either way, although 1000 schools were originally approached only 10% agreed to become involved in the Best Practice study and there was further substantial sample attrition over time.  


The researchers looked at actual levels of individual pupil attainment in their Key Stage 2 (KS2) maths results and compared these with their teacher-led allocation to different maths sets in their subsequent first year of secondary school.

Using KS2 maths data and what we already knew about the group sizes in each school, we made our own decisions about which set each pupil would be in if they were allocated just by KS2 attainment. We then compared our allocations with pupils’ actual set placements“.

All things being equal, therefore, it should follow that two pupils who had the same KS2 score would be placed in the same maths set in their secondary school.



One-third (31%) of pupils were placed in a different set to the one that their prior attainment would predict.

Class (socioeconomic status)

There was no evidence of systematic misallocation based on socioeconomic status (SES). While this fails to replicate the findings of previous research – see, for example, Muijs and Dunne‘s (2010) – it doesn’t automatically follow that class-based misallocation doesn’t occur in the wider school system:

SES differences in attainment at KS2 are the prime driver of inequalities relating to set allocation. In other words, inequalities are already embedded in the education system at age 11 and secondary schooling simply duplicates this already-existing inequality.

The key point from the present study is simply that no evidence has been found that secondary schools exacerbate these patterns of inequality further through their setting processes“.

A further important observation here is that while there were setting misallocations on the basis of SES made by schools, these tended to cancel each other out:

Those from lower socioeconomic backgrounds were both more likely to be misallocated downwards and also more likely to be misallocated upwards“.

The authors explain this in two possible ways:

1. Downward misallocation based on the stereotyping and labelling of working-class students as having “less natural ability”.

2. Upward misallocation based on a “deserving scholarship” imperative. That is, borderline working class children were seen as being “more-deserving” of a chance to work in a higher set.


  • Male students were 1.32 times more likely to be allocated to a higher maths set.
  • Female students were 1.53 times more likely to be allocated to a lower maths set.
  • Ethnicity

  • White students were 2.09 times more likely than Black students to be allocated to a higher maths set.
  • White students were 1.72 times more likely than Asian students to be allocated to a higher maths set.
  • Black students were 2.43 times more likely than White students to be allocated to a lower maths set.
  • Asian students were 1.65 times more likely than White students to be allocated to a lower maths set.
  • Conclusions

    1. The study is the first to demonstrate “there is evidence that setting in Year 7 exacerbates existing educational inequalities in relation to gender and ethnicity“.

    2. The study supports the idea “students should be allocated to sets solely on the basis of prior attainment to avoid misallocation and the ‘creeping prejudice’ implied by our misallocated cases“.

    The study and its findings are interesting for what they show about the process of set allocation in English secondary schools and the implications that it may be substantially infected by a range of prejudices and miscalculations in relation to something that should, according to advocates of setting, be a fairly neutral allocation process based on “measured ability”.

    If you want to take this research further with your students you might want to get them to explore both the nature and extent of the biases involved and their implications: is setting, for example, a simple, objective, exercise based on “measured abilities” or is it something else? And if so, what?

    There are also, of course, wider educational questions to consider: if there are widespread forms of set misallocation in secondary schooling, how is this likely to impact on – and help to explain – differential educational achievement?


    Paul Connolly, Becky Taylor, Becky Francis, Louise Archer, Jeremy Hodgen, Anna Mazenod and Antonina Tereshschenko (2019) “The misallocation of students to academic sets in maths: A study of secondary schools in England“.

    Danial Muijs and Mairead Dunne (2010) “Setting by ability – or is it? A quantitative study of determinants of set placement in English secondary schools

    %d bloggers like this: