BackHome
Refresh: OffView Personal Schedule
Session Submission Summary
Share...

Direct link:

Pathways and architecture to scale: Learning from programs working at scale to improve learning outcomes

Wed, March 13, 8:00 to 9:30am, Hyatt Regency Miami, Floor: Third Level, Pearson 2

Group Submission Type: Formal Panel Session

Proposal

There is a wide consensus that the world is facing a learning crisis (UNESCO 2013, 2015, 2018; World Bank, 2018). To meet the targets of Sustainable Development Goals number 4 (SDG4), governments in low- and middle-income countries are facing increasing pressure to urgently improve the quality of their education systems. This has motivated an increased interest in understanding how to most effectively and quickly scale successful programs that successfully improve student learning. However, most of the evidence available on what works to improve learning outcomes comes from programs operating at small scale (Stern et al. 2021). There is a lack of robust evidence on what works well at scale and why. This is partly due to the fact that it challenging to conduct rigorous research on the causal impacts of system reform (Gibbs et al., 2021).

It is an unfortunate reality that program working at small scale (for example in a few hundred schools) often fail when taken at larger scale (for example, in tens of thousands of schools or more) because they were not designed to work at larger scale or to be delivered through government delivery chains (Gibbs et al, 2021). It is promising that recent research has started to uncover some of the common ingredients at the classroom level and at the system level that are associated with successful programs operating at large scale (Stern et al. 2021). In particular, the importance of aligning with government priorities, designing within governments’ existing systems, and leveraging / building government capacity at the middle-tier level are key recurrent factors of success.

As there are pockets of excellence in any system, there are examples of programs that have successfully been operating at scale to drive improvement in learning outcomes. This Formal Panel Session will bring together lessons from scale from Botswana, Zambia and India to uncover some of the key learnings they have experienced in scaling education programs within government systems. These case studies will reflect on the levers to scaling that were unlocked in different interventions (for example, fostering a supportive environment for teachers and empowering them to harness data, establishing positive communication channels, distributing decision making by engaging sub-national education leadership, creating cultures that enable adaptation and innovation, and responding to demand at different levels of the system.)

In particular, the Formal Panel Session will share lessons on:
1. How these different programs designed their intervention from the start to make it simple enough to work in the government system.
2. The marginal additional cost per student of the intervention and the key drivers for this.
3. How the different programs engage with the middle layers (i.e., stakeholders at different levels of the government such as headteachers, district officers, managers of coaches) to ensure that the schools lagging behind receive the right support and to sustain interventions over time.
4. The types and level of political engagement at the central level that was necessary to sustain momentum.
5. Specific issues encountered when working through the Government system (e g., books arriving on time etc) and how these were mitigated.

The aim of this panel will be to present evidence of elements that worked well in specific national contexts in order to contribute to a wider discussion to be facilitated by the Discussant, on how these learnings could be applied to other national contexts and how they might lead to further evidence generation efforts to continue building the evidence-base in this emerging sector of research at scale.

Sub Unit

Organizer

Chair

Individual Presentations

Discussant