Abstract | The National Tutoring Programme (NTP) Tuition Partners (TP) programme was designed to provide additional support to schools and teachers to supplement classroom teaching through subsidised high-quality tutoring for pupils from an approved list of tutoring organisations, the Tuition Partners. This evaluation covers the TP programme as delivered in its first year by the Education Endowment Foundation (EEF), from November 2020 to August 2021. Tuition Partners was one arm of the NTP. The NTP aimed to support teachers and schools in providing a sustained response to the Covid-19 pandemic and to provide a longer term contribution to closing the attainment gap between disadvantaged pupils and their peers. The NTP was part of a wider government response to the pandemic, funded by the Department for Education and originally developed by the EEF, Nesta, Impetus, The Sutton Trust, and Teach First, and with the support of the KPMG Foundation. The EEF appointed 33 approved ‘Tuition Partners’ that schools could select from to deliver tuition. Schools could access 15 hours of tutoring per selected pupil (with a minimum of 12 hours being considered a completed block of tuition). Tuition was provided online and/or face-to-face; and was 1:1, or in small groups (1:2 or 1:3); and available in English, maths, science, humanities and modern foreign languages. Tuition was expected to be delivered in schools (before, during and after school), in addition to usual teaching; and, in certain circumstances, at home. The programme was targeted at disadvantaged pupils attending state-maintained schools in England, including those eligible for Pupil Premium funding (PP-eligible), Free School Meals (FSM), or those identified by schools as having an equivalent need for support. Participating schools had discretion to identify which of their pupils they felt would most benefit from additional tuition support. Pupils in Years 1–11 were eligible (5–16 years old). The programme aimed to reach 215,000 to 265,000 pupils, across 6000 state-maintained schools in England, and it was expected that approximately 20,000 tutors would be recruited by Tuition Partners. The TP programme was set up and delivered during the Covid-19 pandemic, requiring continued responsiveness to the challenges faced by schools including restricted attendance, remote teaching, and ongoing widespread staff and pupil absences. During the school closures to most pupils from January – March 2021, the EEF approved TPs to deliver online tuition at home, however many schools chose to wait to commence tutoring until schools reopened fully, and therefore started tutoring later than planned. This evaluation report covers the analysis on the impact of the TP programme on the maths and English attainment outcomes for primary school pupils (Years 1–6) using standardised classroom assessments. Separate reports relate to analysis on Year 11 pupils and an implementation and process evaluation (IPE). The evaluation findings for the TP programme are brought together in a summary and interpretation report that is available here. This evaluation uses a quasi-experimental design (QED), involving a group of intervention schools that participated in the TP programme, and a group of comparison schools that did not receive the programme. The evaluation relies on a propensity score matching and re-weighting approach to ensure that the intervention and comparison schools are similar to each other in important, observable regards. As pupils who would have received TP in comparison schools were difficult to identify, the evaluation focused on pupils eligible for Pupil Premium and on all pupils, as these groups can be identified in both TP and comparison schools. For English, the analysis is based on 165 primary schools with 7073 pupils eligible for Pupil Premium and for maths, 126 primary schools with 5102 pupils eligible for Pupil Premium3. An additional instrumental variable (IV) analysis, based on the sample of TP schools only, looked at the impact of TP in schools that signed up to the TP programme earlier (and that delivered more tutoring) compared to schools that signed up later. On average, pupils eligible for Pupil Premium in schools that received TP made similar progress in English and maths compared to pupils eligible for Pupil Premium in comparison schools (no evidence of an effect in English or in maths). This result has a low security rating. A particular challenge is that, on average, only approximately 20% of pupils eligible for Pupil Premium were selected for tutoring, meaning a large proportion of pupils eligible for Pupil Premium were included in the analysis who did not receive tutoring. Therefore, this estimated impact of TP is diluted and it is hard to detect any effect that may (or may not) be present. Similar analysis on all pupils found that pupils in schools that received TP made, on average, similar progress in English compared to all pupils in comparison schools (no evidence of an effect), and an additional one month’s progress in maths compared to pupils in comparison schools. However, there is uncertainty around these estimates, with the positive maths result being consistent with a null (0 months) or slightly larger positive effect (2 months) and the English result being consistent with small positive (1 month) or small negative effect (−1 months). Furthermore, this analysis was subject to even further dilution: on average, only 12% (for maths) and 14% (for English) of pupils in the analysed schools were selected for tutoring. Given this context, it is unlikely that any of these differences were due to TP. In the sample of TP schools, completing a 12-hour block of tutoring (compared to zero hours) was related to higher English scores amongst pupils eligible for Pupil Premium that received more tutoring due to the early sign-up of the school. An equivalent analysis for maths was not able to proceed. A different analysis within TP schools showed that pupils who received more hours of tutoring were associated with higher English scores on average than pupils who received fewer hours of tutoring. However, this was not the case for maths, where receiving more hours of tutoring was not associated with higher maths scores. These results are associations and are not necessarily causal estimates of impact; there may be other explanations for the results. |
---|