Same Day Intervention (SDI) is a mathematics intervention developed and delivered by a partnership between Yorkshire and the Humber Maths Hub and Outwood Institute of Education. In this trial it was delivered to Year 5 pupils (age 9 – 10).
Usual mathematics lessons are replaced with 75-minute SDI lessons, which are separated into three parts. In part one, teachers teach a mathematics concept before giving pupils questions to assess their understanding. In part two, teachers then mark pupil responses, while pupils attend an assembly or are taught by a Teaching Assistant (TA). In the final part, the teacher then uses the assessment to separate the class into two, providing additional instruction to pupils who performed less well in the test, while a TA supports the rest of the class. In this trial, SDI mathematics lessons were delivered daily for 7 months.
The EEF’s Improving Mathematics in Key Stages Two and Three: Evidence Review explains that mastery approaches in maths are supported by some evidence (Hodgen et al 2018). However, this evidence is drawn largely from US studies, and further research is required into the efficacy of mastery-based approaches in English schools, which are becoming increasingly popular. SDI features components of mastery learning (such as frequent feedback, and extra time for pupils who struggle), while also drawing upon other promising approaches identified by the EEF review (such as feedback and formative assessment). Yorkshire and the Humber Maths Hub and Outwood Institute of Education aimed to use the approach to improve maths attainment for all pupils, while also narrowing the attainment gap between lower and higher attaining pupils. Given that the intervention requires teachers to mark in lesson, it was also hypothesised that SDI may reduce teacher workload, with teachers having less to mark outside of the classroom.
In this trial, pupils in Same Day Intervention (SDI) schools made, on average, 0 additional months’ progress compared to those in the control equivalent, and this result had a moderate to high security rating. Exploratory analysis finds the same result for children eligible for Free School meals
Half of teachers in SDI schools provided data on the time they spent marking pupil work, and there was evidence they spent less time marking than teachers in control schools (approximately one hour less per week). However, given that data is only available for half of teachers, this should be treated with caution. Findings from the Implementation and Process Evaluation also suggested that marking time may have decreased, however, it also indicated that the time taken to plan lessons may have increased. Teachers in SDI schools were more likely than teachers in control schools to report that all pupils, and lower achieving pupils, were more confident compared to the previous year’s pupils. It should be noted that this is the perception of teachers, rather than pupils. Pupil focus groups in case study schools suggested that pupils had mixed feelings towards SDI
Less than half of intervention schools implemented SDI as intended, and schools faced several challenges which made effective implementation difficult. For instance, schools were required to ensure a TA was present in every Year 5 mathematics lesson, as their support was a core component of the intervention. The post-intervention teacher survey found that 25% of schools always had a TA present, 46% regularly did, 18% only occasionally did, while 9% never did. Schools also found it challenging to adapt the school day so that sessions could last the intended 75 minutes. Other aspects of delivery (such as the approach to modelling concepts, pupil assessment questions and the division of the class into targeted groups) were often not implemented with fidelity, while attendance at training was also mixed. Only 55% of teachers attended all 3 compulsory training sessions.
Poor implementation may have contributed to the overall lack of impact. Compliance analysis suggests that when SDI was implemented with greater fidelity, pupils made, on average, two +2months’ additional progress in mathematics (with the possible range of effects varying from no additional progress to four +4months’ additional progress). Other factors may also have contributed to the lack of impact, including the programme not being taught for long enough, or control schools using practices that were similar to those introduced by SDI.
The EEF has no plans to fund a further trial of Same Day Intervention, but continues to be interested in mastery approaches in mathematics.
- Pupils in Same Day Intervention schools made the equivalent of zero additional months’ progress in mathematics, on average, compared to pupils in control schools. This result has a moderate to high security rating.
- Half of teachers in Same Day Intervention schools provided data on the time they spent marking pupil work, and there was evidence that they spent less time marking than teachers in control schools (approximately one hour less per week). Teachers in Same Day Intervention schools were also more likely than teachers in control schools to report that all pupils, and lower achieving pupils, were more confident compared to the previous year’s pupils.
- Exploratory analysis suggests that pupils eligible for Free School Meals (FSM) in Same Day Intervention schools made the equivalent of zero months’ progress in mathematics, on average, compared to FSM pupils in control schools.
- Fewer than half of intervention schools implemented Same Day Intervention as intended. Schools faced several challenges which made implementation difficult, such as being able to provide a teaching assistant to support delivery and adapting the school day to extend mathematics lessons.
- Compliance analysis suggests that pupils in schools that implemented Same Day Intervention as the programme developer intended may have made, on average, two months additional progress in mathematics compared to pupils in control schools. There may be other differences in schools that implemented the programme with greater fidelity, however, making this finding less secure than the primary analysis.