Student allocation to maths sets not always based on previous achievement

Research into grouping by achievement, by academics from Queen’s University Belfast and University College London, has found that nearly a third of students in England were allocated to higher or lower maths sets than their previous test performance implied.

The study, published in the British Educational Research Journal, analysed data from 9,301 Year 7 students at 46 secondary schools in England. The researchers compared which maths set the students would have been put in – based on Key Stage 2 maths test scores – with the sets they were actually placed in. Overall, they found that 31.1% of students were misallocated – placed in sets that were either higher or lower than their results at the end of primary school would have indicated.

Boys were slightly more likely to be misallocated to higher sets in maths (16.7%) than lower sets (13.0%), whereas girls were more likely to be misallocated to lower sets (17.9%) than higher sets (14.7%). Other findings showed that:

  • Black students were 2.4 times more likely than white students to be misallocated to a lower maths set.
  • Asian students were 1.7 times more likely than white students to be misallocated to a lower maths set.
  • Female students were 1.53 times more likely than males to be misallocated to a lower maths set.
  • White students were 2.09 times more likely than black students to be misallocated to a higher maths set.
  • White students were 1.72 times more likely than Asian students to be misallocated to a higher maths set.
  • Male students were 1.32 times more likely than females to be misallocated to a higher maths set.

Source: The misallocation of students to academic sets in maths: A study of secondary schools in England (June 2019) British Educational Research Journal

Grouping pupils by achievement

The Education Endowment Foundation has published an evaluation of two trials of programmes developed by the University College-London (UCL) Institute of Education investigating approaches to grouping pupils: Best Practice in Setting and Best Practice in Mixed Attainment Grouping.

The main trial, “Best Practice in Setting”, tested an intervention that aimed to get schools to improve their setting practice (grouping pupils in classes by their current achievement levels). A total of 127 schools took part in the trial, which ran over the course of two academic years. Teachers were randomly allocated to sets to prevent “lower” sets from being disproportionately assigned less-experienced teachers, while pupils in Years 7 and 8 were assigned to sets based on independent measures of achievement, rather than more subjective judgements such as behaviour and peer interactions. There were opportunities throughout the year to re-assign pupils to different sets based on their current level of achievement.

The evaluation found no evidence that the intervention improves outcomes in maths (effect size = -0.01) or English (effect size = -0.08). The process evaluation revealed mixed views from participants, and many interviewees thought that what they were being asked to do represented little change from what they already do.

The researchers noted that because school and teacher buy-in was low and attrition rates for follow-up testing were high, half of the schools in the math trial and more than half of the schools in the English trial stopped the intervention before follow-up, and this makes it difficult to conclude anything certain about the impact of Best Practice in Setting.

Source: Best practice in grouping students. Intervention A: Best practice in setting evaluation report and executive summary, (September 2018). Education Endowment Foundation

Best practice in grouping students. Intervention B: Mixed attainment grouping. Pilot report and executive summary, (September 2018). Education Endowment Foundation

Why do so few low- and middle-income children attend a grammar school?

A new study led by John Jerrim at UCL Institute of Education suggests that private tutoring may be one reason that children from high-income families are more likely to get into grammar schools than children from low-income families.

The research, which was funded by the Nuffield Foundation, uses data from the Millennium Cohort Study for more than 1,800 children from grammar school areas in England and Northern Ireland. It considers how factors such as family income, prior academic achievement, private tutoring and parental attitudes and aspirations are linked with children’s chances of attending a grammar school.

The study finds that children from families in the bottom quarter of household incomes in England have less than a 10% chance of attending a grammar school. This compares to around a 40% chance for children in the top quarter of household incomes. Results also show that children who receive tutoring to prepare for grammar school entrance exams are more likely to get in. Overall, around 70% of those who receive tutoring get into a grammar school, compared to just 14% of those who do not. However, less than 10% of children from families with below average income receive tutoring for the grammar school entrance test, compared with around 30% of children from households in the top quarter of family incomes.

Source: Why do so few low and middle-income children attend a grammar school?  New evidence from the Millennium Cohort Study (March 2018), UCL Institute of Education