Switch-on Reading (re-grant)

This page covers the second (effectiveness) trial of Switch-on Reading, testing a scalable model under everyday conditions in a large number of schools. To read about the first (efficacy) trial - testing whether it could work in schools under best possible conditions - click here.

Switch-on is an intensive, targeted literacy intervention that aims to improve the reading skills of pupils who are struggling with literacy. There are two versions of the intervention: Switch-on Reading and Switch-on Reading and Writing. Both involve specially trained Teaching Assistants (TAs) delivering a tailored programme of literacy support in daily 20-minute sessions over a ten-week period.

A previous EEF-funded evaluation of Switch-on had shown signs of promise in raising reading outcomes for Year 7 pupils, hence a larger scale effectiveness evaluation was conducted. Each of the 184 participating schools were randomly assigned to receive either Switch-on Reading, Switch-on Reading and Writing, or continue their usual practices of supporting pupils with reading difficulties (called ‘business as usual’ control). Between two and four TAs in each school were trained to deliver the interventions by a team of trainers who were distinct from the developers. TAs received 1.5 days training for Switch-on Reading, and 2.5 days for Switch-on Reading and Writing. Each TA worked with between two and four Year 3 pupils who had not met age-related expectations in literacy at the end of Key Stage 1, and who did not have a high level of special needs. In total, 999 pupils were involved in the trial. The primary outcome measure was the Hodder Group Reading Test 2A. The evaluation was conducted over the 2015/2016 academic year. A process and implementation evaluation was carried out which included depth telephone interviews with the developers, trainers, and TAs, and six case studies incorporating the views of TAs, class teachers, and Switch-on co-ordinators in the school.

Key Conclusions

The following conclusions summarise the project outcome

  1. Participating children in schools delivering either version of Switch-on made no additional progress in reading compared to similarly struggling children in ‘business as usual’ control schools. The 4 padlock security rating means that we have high confidence that there was no difference, and that this was due to Switch-on and not affected by other factors.

  2. A similar result was observed for children eligible for Free School Meals.

  3. The secondary analysis suggests that Switch-on might have affected children who did not actually receive the intervention by changing the make-up of their class or the capacity of their TAs. These findings are tentative, but emphasise the importance of considering potential impacts on all children when using targeted interventions.

  4. Overall, participating staff were positive about the intervention and accompanying training. Perceived outcomes for pupils included increased confidence, motivation and interest in reading, and improved reading and writing skills. Skills of participating TAs were also felt to have improved.

  5. Some schools reported modifications to the prescribed content, duration and format of Switch-on sessions. Closer monitoring by the developers could help to ensure greater fidelity in future. It was also reported that some class teachers had limited awareness of Switch-on. Clarifying the role of Switch-on as part of a broader, teacher-led literacy strategy could improve implementation.

What is the impact?

The trial did not find evidence that a three-month delivery of Switch-on Reading or Switch-on Reading and Writing improves reading outcomes of pupils struggling with literacy at Key Stage 1 compared to schools’ usual practices. Estimated effects were at zero standard deviations and not statistically significant. The intervention also showed no effect on pupils eligible for free school meals. These findings contradict previous evidence from a smaller within-school efficacy trial and the literature on the effectiveness of structured, one-on-one literacy interventions led by TAs.

The process evaluation suggests that in some schools intervention fidelity was compromised to some degree, which is likely to have affected the quality of the intervention and could explain the lack of impact. The process evaluation also indicated inconsistent involvement of class teachers and senior leaders, which is identified in the literature as a factor that enables effective implementation.

The secondary analysis suggests that Switch-on might have affected children who didn’t actually receive the intervention by changing the make-up of their class or the capacity of their TAs. There is a possibility that implementing Switch-on may have had a negative impact on the literacy of pupils in the same year who were eligible to receive the intervention but were not selected due to limitations on numbers. Conversely, there is a possibility that the intervention had a positive impact for pupils in the same that class that were ineligible to receive the intervention. Both results should be treated cautiously.

GroupEffect size (95% CI)Months' ProgressEEF Security RatingType of TrialEEF Cost Rating
Switch-on vs. control 0.00 (-0.13, 0.13)0 months
Switch-on vs. control FSM0.03 (-0.18, 0.20)0 Monthsn/aEffectiveness

How secure is the finding?

The security rating of the trial indicates how confident we can be that any additional progress experienced by the children receiving Switch-on was due to the intervention rather than any other factors. This trial was an effectiveness trial that aimed to test whether the intervention was effective when delivered in a form that could be available to a large number of schools. Findings from this study have high security. The trial was a well-designed, three-arm randomised controlled trial, with randomisation at school level. Relatively few pupils were lost to the analysis and pupils who received the intervention were similar to the pupils in the comparison group. The trial was only designed to reliably detect an impact equivalent to three months’ progress.

How much does it cost?

The cost of the project was estimated at £546 per pupil. This estimate includes up-front costs (for website development and access, and for the development of training materials such as films) and ongoing delivery costs (including provision of books and TA and trainer training). In terms of staff time, approximately 68 hours per pupil were required to deliver Switch-on. This included training and the delivery of sessions. If the intervention were delivered in a school over a period of three years, the cost per pupil per year would be much lower, at £184, because most of the cost are up-front costs. Estimates are based on an average of five pupils per school and two or three TAs per school delivering the intervention.

EEF summary

In a previous EEF trial, Switch-on Reading was found to deliver around 3 months additional progress in reading outcomes in Year 7. In that trial, the Switch-on training was delivered by its original developers. This new project was designed to test whether Switch-on Reading (and Switch on Reading and Writing) would have an impact using the type of delivery model that would be needed to make it available to a large number of schools, without direct developer involvement.

In this second trial, participating children in Switch-on schools made no additional progress in reading compared to similarly struggling children in ‘business as usual’ control schools. The difference in the two results could be because the delivery approach was changed, or because the trials were done with pupils of different ages. Alternatively, the difference could be due to changes in the evaluation process. For example, the amount of other literacy provision available to pupils in the comparison group appears to have been higher in the second trial.

Because of the positive result from the first trial, EEF is discussing with the Switch-on team whether it is possible to develop and test a model that is deliverable in a large number of schools while retaining more of the key elements of the original intervention. Schools considering using Switch-on now should aim to make the conditions as similar as possible to those in that first trial. Involvement from the original developers could be beneficial.