Further Education Attendance Strategies – Setting Choices

This study is a ​‘Setting Choices’ project, that aims to explore what strategies are being used by Further Education (FE) colleges to address poor attendance, including best practices and how they may lead to improvements in learners’ attendance and attainment outcomes.

16 – 19 setting leaders make choices about setting-wide practices and approaches that are intended to produce positive outcomes for learners, such as how to support learners as they transition to college, or how to provide targeted supports for socioeconomically disadvantaged learners. The aim of Setting Choices research is to produce causal evidence about the impact of different setting-level approaches and policies on outcomes of interest, with particular attention to impact on learners from disadvantaged backgrounds.

Setting Choices evaluations are divided into two phases: a scoping phase, and an impact evaluation phase. During the scoping phase, the research teams will refine their research questions, verify their assumptions, and explore the feasibility of their evaluation designs.

Attendance is a significant challenge in 16 – 19 education, with average attendance rates lower in FE colleges than in secondary schools, especially in English and maths. The challenge is greater for disadvantaged learners, who are more likely to be in paid work or face other limitations due to financial difficulties. Research has shown that persistent absences have a negative impact on maths and reading attainment, and that failure to achieve a Level 4 in GCSE maths and/​or English affects learners’ future life chances (College Attendance Survey Report, Association of Colleges, 2024).

FE colleges use a range of strategies to encourage attendance, which can be targeted at low-attendance learners or the general learner population. There is little evidence supporting these strategies or whether they improve learners’ attendance and attainment outcomes. Furthermore, while FE colleges track learner attendance granularly in their own management information systems (MIS), the lack of nationally agreed definitions and formal data sharing arrangements are key barriers to understanding what strategies and data collection systems are most effective in supporting learner engagement and other outcomes.

This project aims to address these evidence gaps by building our understanding of the strategies being used by FE colleges to address poor attendance and of the attendance data landscape. The project’s findings around the data landscape may also be useful for future 16 – 19 research and to those interested in benchmarking.

This project is currently at the scoping phase. The aim of this phase is to identify what strategies FE colleges are using to improve attendance, what data on learners’ attendance and potential secondary outcomes FE colleges are collecting and how, to assess the feasibility of, and recommend the preferred approach for, robustly evaluating the impact of FE colleges’ attendance strategy choices.

The scoping phase will involve:

• Identifying one or more attendance strategies that are relevant to FE colleges and have a potential for impact on attendance and attainment outcomes, with an emphasis on disadvantaged learners.
• Developing a theory of change for the attendance strategy(ies).
• Mapping attendance data systems and determining whether it is possible to collect attendance data from and consolidate it across FE colleges to define appropriate outcome measures.
• Determining whether an impact evaluation of the attendance strategy(ies) is feasible, and whether a randomised controlled trialAn RCT is used evaluate an educational programme by assigning settings to one of two groups: the intervention group, who receive the programme or the control group, who continue with business as usual. This ensures that any differences in outcomes can be confidently attributed to the programme, providing a robust estimate of the impact and contributing to the evidence for what works in improving educational outcomes. or quasi-experimental design should be the preferred approach.