Optimising Messaging for Returns to School

Introduction

The use of messaging in education is listed as the only ‘great buy’ in the World Bank’s comparative analysis of cost-effectiveness in EdTech interventions. But, the different studies on messaging cited in the report represent different approaches and contexts. One analysis of nudge messaging in education suggests that interventions often provide diverging results, with outcomes dependent on context. Two questions remain:  

• “Why does messaging work in some contexts, and not in others?”
• “What is it about messaging that works?”

Within the broader landscape of messaging research in education, nudges to improve attendance, limiting learning loss during Covid-19, and learning applications that use ‘messaging’ as the medium are often bundled together. It is no surprise that these approaches have different impacts in different contexts. There is also a temptation to ignore or not publish findings that don’t fit this pattern, and view this low-cost approach as a silver bullet to accessing education.

EdTech Hub has recently commissioned several studies which aim to add to this contextual understanding. Through exploring the fundamentals of messaging and better understanding the circumstances in which it is used, EdTech Hub strives to bridge this gap in understanding what works in messaging and why. 

EdTech Hub study on optimising messaging for returns to school in Ghana

Among the EdTech Hub studies addressing this issue, our study titled ‘Optimising messaging to promote returns to school in Ghana for marginalised learners’, will explore the main characteristics of messages — to determine the impact that changing these characteristics has on students attendance rate. The study will therefore provide answers on how messages themselves can be adapted and optimised to most effectively promote returns to education among marginalised groups in Ghana.

This framing will allow us to explore the currently unanswered questions on the contextual factors that influence why messaging interventions work in some contexts, and for some groups, and not for others. It will also help to address the issue of low returns to school post-Covid, particularly among girls and marginalised groups.

For this research, we will be working in rural Ghana in collaboration with Craft Education Technology as the technical partner. The study will use their platform that uses messaging technology to deliver content to learners in rural and marginalised communities. Throughout the study, the arrangement, content, and language of messages will be iterated, to determine the impact that tailoring these characteristics has on participation in school.

How does this further the conversation on messaging?

A more critical and nuanced approach to messaging informs evidence-based decisions on deploying messaging at scale. This study, in combination with other commissioned research on participation and messaging within the Hub, will provide evidence on how messaging can be used in different contexts to increase its impact on participation in education. 

Interventions aiming to improve participation should also be linked to evidence on how that participation leads to improved learning outcomes. This study engages with this by measuring sustained engagement with messages, and its link to learning progression. The role of sustained engagement in increased participation and maintaining a learner identity can be a particularly valuable contribution for marginalised learners whose learner identity may be undermined by others.

In the current context of reduced participation in education due to Covid-19, the fundamental topic is how evidence-based decisions on messaging can be made to re-engage students in education, reduce the number of out-of-school children, and improve learning outcomes over time.  This study will also provide evidence for a broader policy shift to address wrong assumptions of messaging interventions being inevitably and uncritically effective.

How can we make messaging work?

This study, therefore, addresses an imperative for evidence on messaging to engage critically. Alongside this, the study tackles the need for research on messaging to adapt to the needs of different marginalised groups; a need which has been highlighted by recent research. A working paper showed that the Covid-19 pandemic disproportionately reduced learning for the most economically marginalised students, who scored significantly lower in literacy and numeracy tests than their peers. A forthcoming study in collaboration with Innovations for Poverty Action (IPA) also demonstrated that SMS nudges delivered varied results, increasing expectation on student performance (particularly for girls) among caregivers with no education, which subsequently reduced these students’ engagement and participation. This underlines the appropriateness and importance of this study adopting a critical approach to messaging and differentiating between marginalised groups.

Through building on existing research and adopting a focus on message characteristics, this study will help answer the critical question of what aspects of messaging are most effective in increasing participation in education for marginalised groups. This will contribute evidence that is essential to develop messaging interventions that can be used effectively in different contexts, for different audiences, to provide meaningful long-term impact on learning outcomes. The ambition is that the study can inform future research and policy, to enable messaging interventions in different contexts to be optimised for cost-effective learning gains in a range of contexts.