Best Practice in Setting

Best Practice in Setting aims to improve the educational attainment and self-confidence of students who are currently placed in attainment groups for maths or English, by preventing poor setting practices. Teachers were trained in how to use best practice principles in their schools through four twilight training workshops. 

Another evaluation piloted mixed attainment teaching practices.


Key Stage 3

Key stage


Cross curriculum


EEF Summary

Many schools group students by attainment, particularly in English and Maths, but the evidence suggests that setting by class does not on average have a positive impact. This project aimed to see if outcomes for students could be improved if some of the potentially negative aspects of setting were addressed.

The study provides no evidence that the Best Practice in Setting intervention had a positive impact on maths or English outcomes. However, the results have low to moderate security for the maths finding and very low security for the English finding because a large number of participant schools dropped out of the programme.

Attendance at workshops and adherence to the some of the best practice principles was low. Because of this, we cannot conclude that the underlying idea – improving setting practices to tackle poorer outcomes for those in lower sets – does not work. However, it demonstrates that this particular programme was not effective at supporting the schools to adopt new setting practices.

Setting is an important issue for schools and EEF would like to generate more evidence in this area. It is clear from this trial that it is challenging for schools to adapt their setting practices and careful thought needs to be given to how best we can generate useful evidence in this area. 

Research Results


Months' Progress
Evidence Strength


Months' Progress
Evidence Strength

Maths (everFSM)

Months' Progress


English (everFSM)

Months' Progress


Were the schools in the trial similar to my school?

There were 127 schools participating in the trial. In the maths trial there were 121 schools, whilst in the English trial there were 79 schools, with 73 schools participating in both trials. The majority of schools in both trials were in urban locations and 82% and 78% of schools were Good or Outstanding in the maths and English trials respectively. 

Could I implement this in my school?

The programme is not currently available for schools to purchase. Schools interested in the approach could work to implement the best practice principles without external support, however caution should be taken as there is no evidence that this will result in an improvement to pupil outcomes and schools may find the approaches difficult to implement. 



Delivered by


Whole Class

Participant group


2 Years

Intervention length

How much will it cost?

The average cost of the Best Practice in Setting intervention per department was around £2,500, or £14.04 per pupil per year when averaged over three years. The main costs were the cost of teacher training and programme materials. Schools also needed to meet the cost of staff cover for two days of teacher training per department for the first two years of the intervention.



Cost per pupil


All Teachers

No. of Teachers/TAs


4 Days

Training time per staff member

Evaluation info





Key Stage

Key Stage 3

Start date

January 2014

End date

September 2018

Type of trial

Efficacy Trial

Evaluation Conclusions

  1. The project found no evidence that the Best Practice in Setting intervention improves maths or English attainment for children in Years 7 and 8. For English, pupils in the intervention schools made slightly less progress than the control pupils, but this finding has no meaning given its level of uncertainty and very low security rating.

  2. There was no evidence of impact on pupils’ self-confidence in maths. For English there was a small positive difference in pupil self-confidence, but this result has no meaning due to its level of uncertainty and the large amount of measurement attrition.

  3. School and teacher buy-in was low. Half of the schools in the maths trial and more than half of the schools in the English trial ceased intervention delivery before follow-up. Attendance at training sessions decreased over time (to 21% for the final maths session and 12% for the final English session). With this level of treatment attrition, being able to engage schools with the programme was demonstrated to be at least as important as the programme itself. Future interventions aimed at changing setting practices should be designed with the issue of engagement in mind. 

  4. The process evaluation revealed mixed views from participants. While some were largely positive about the intervention, many interviewees thought that what they were being asked to do represented little change from what they already do. Some schools reported that the intervention was onerous and difficult to deliver. 

  5. Some schools struggled to implement the intervention in full (for example, there was poor fidelity to teachers being randomly allocated to sets), this seemed to be due to schools finding it difficult to make some changes and therefore adapting principles to make them more implementable. 

  1. Updated: 6th September, 2018

    Printable project summary

    1 MB pdf - EEF-best-practice-in-grouping-students.pdf

  2. Updated: 6th September, 2018

    Evaluation report

    1 MB pdf - Intervention_A_-_Best_Practice_in_Setting.pdf

  3. Updated: 12th June, 2017

    Intervention A Evaluation Protocol

    399 KB pdf - Round_6-_Intervention_A_Best_Practice_in_Setting_Amended_version.pdf

  4. Updated: 7th November, 2017

    Intervention B Statistical Analysis Plan

    370 KB pdf - Round_6-_Ability_grouping_BPS_SAP.pdf

Full project description

The Best Practice in Setting intervention was designed by academics at UCL Institute of Education to improve the educational attainment and self-confidence of students in Years 7 and 8 who are currently placed in attainment groups for maths and/or English.

Setting (which we use to mean grouping students in classes by their current attainment levels) is practiced in England to a varying degree and the extent of the practice can be dependent upon subject, year group, or resource availability (Kutnick et al., 2005). However, by Key Stage 3, setting is commonplace in mathematics (Ofsted, 2013) and is also increasingly introduced in English in the later stages of Key Stage 3 and as students enter Key Stage 4 (Ofsted, 2013).

Setting or streaming appears to benefit higher attaining pupils whilst being detrimental to the learning of lower attaining learners (see the Sutton Trust-EEF Teaching and Learning Toolkit; Slavin, 1990; and Sukhnandan and Lee, 1998). There are a number of practices associated with setting that may be the cause of this differential effect. This intervention was designed to raise pupil attainment by helping schools address these previously-identified poor setting practices through the application of best practice principles, including:

  •  assigning teachers randomly to sets to ensure all pupils received high quality teaching;
  • focusing on pedagogy (through training and materials);
  • having no more than four different sets for any subject, which resulted in broader sets and aimed to reduce hierarchies and any negative impacts of misallocation;
  • assigning students to sets based on independent measures of prior attainment, rather than subjective teacher assessment; and
  • providing the opportunity to reassign students to different sets as suggested by their attainment results over the year. 

Teachers were trained in how to use the principles in their schools and then expected to apply them. The training was delivered by UCL Institute of Education researchers over the course of two academic years (2015/2016 and 2016/2017) and performance in English and maths tests was measured for a cohort of students that progressed from Year 7 to Year 8 during this time. The evaluation was set up as a school-randomised trial. One hundred and twenty seven schools were randomised to either receive the Best Practice in Setting intervention (in English and/or maths) or continue with their existing setting practices.

For the process evaluation, surveys and interview data were used to measure implementation and fidelity, explore organisational and attitudinal changes, and assess scalability.