About
This mixed-methods study explores how the Government of Sierra Leone used a Geographic Information System (GIS)-supported preference matching algorithm to improve teacher allocation, aiming to address stark urban–rural disparities in pupil–to-qualified-teacher ratios and make teacher deployment fairer and more effective. The research qualitatively and quantitatively evaluates the impact of the centralised, tech-enabled teacher deployment process on decision-making, equitable teacher allocation, and achievement of policy goals. Additionally, it examines the practicalities of using a preference matching algorithm and the challenges of implementing a data-driven system in education planning.
Key Findings
The implementation research components of this study embed in Sierra Leone’s government a data-driven approach to teacher allocation, with the aim of improving teacher recruitment, retention and satisfaction. A cutting-edge algorithm for teacher allocation, developed by partner Fab inc, was piloted at a national level for the first time, with embedded training for the Teaching Service Commission. The algorithm is informed by and improved by AI. It matches teachers with schools by applying a set of pre-determined rules and conditions. Findings include:
- The implementation of the algorithm-based teacher deployment in 2024 marked a significant shift toward a more data-driven, transparent process. While the new approach demonstrated clear benefits—including improved fairness, accountability, transparency and reduced political interference —it also revealed important challenges related to data quality, governance, and communication.
- The study illustrated that multi-level government buy-in is crucial to effective data-driven decision-making at scale. The shift to centralised, automated decision-making reduced discretionary control at the local level, leading to resistance from some stakeholders.
The centralised, tech-enabled teacher deployment process, supported by the preference matching algorithm, resulted in the majority of the Teaching Service Commission priorities being met, including all newly added teachers passing the licensing exam, increasing the percentage of qualified teachers, and deploying teachers:
- to school with worse pupil-to-payroll-teacher ratios,
- to more remote schools, and,
- in districts defined as disadvantaged.
Bridging the Gap
The Challenge
The school workforce plays a key role in influencing learning outcomes. However, decision-makers in low-income countries face challenges in achieving an equitable distribution of teachers. In Sierra Leone, this challenge is particularly acute. The pupil–to-qualified-teacher ratio increases from 44:1 in urban centres to 76:1 in rural areas, indicating significant disparities in teacher availability across locations.
Why It Matters
Embedding data-driven learning cycles in ministry decision-making is a significant challenge. Optimising allocation of teachers across a national educational system could have a significant cost-effective impact in improving teacher retention and reducing absenteeism.
How this Work is Aiming to Address It
In 2024, the Government of Sierra Leone used a GIS-supported preference matching algorithm to improve teacher allocation. This mixed-methods study built evidence on the impact of the centralised, tech-enabled teacher deployment process on decision-making, equitable teacher allocation, and achievement of policy goals.
Objective
The research aims to inform data-driven approaches in teacher allocation at a national level as a means of addressing Sierra Leone’s teacher shortage crisis. Attracting and retaining more teachers, and placing more teachers into disadvantaged schools, is vital to improving education in Sierra Leone.
The Research Questions
- How can decision-makers in low- and middle-income countries improve teacher allocation processes?
- What impact does a GIS-supported teacher preference matching algorithm have on teacher deployment decision-making?
Study Design and Methodology
The research had two components:
- The quantitative study presents a pre- / post-analysis of the uptake of 2341 new teacher allocations, and examines the practicalities of implementing a data-driven process, amid data challenges and shifting priorities.
- The qualitative study examines if and how the new deployment process, enabled by the algorithm, affects the government’s decision-making process on teacher allocation, and whether it supported the Teaching Service Commission in its policy goals.
Timeline of Activities
2021
Inception phase
Inception report: The impact of GIS-supported teacher allocation in Sierra Leone
2021 – 2022
Webinars
Webinar series on GIS in education
2022
Work starts
Research study began in January 2022
The Importance of the Results
Collaborating with the Teacher Service Commission from the beginning through Design-Based Implementation Research (DBIR) has meant that this study is embedded practically in the real-world context, and served to strengthen the government’s capacity in data-decision making. The DBIR also showed the importance of robust relationships, especially as the study has been maintained despite an election and delay in new teacher deployment.
The adoption of a data-driven approach to teacher allocation that considers school need, demographics, teacher preferences, and other key factors has the potential to improve teacher retention, attendance, and overall job satisfaction. In turn, this has the potential to strengthen and replenish the teacher workforce in a highly cost-effective way, improving transparency and accountability, while also reducing absenteeism and improving teaching quality. This pilot represents the first time a GIS-support preference matching algorithm for teacher allocation has been implemented at a national level alongside embedded government training in Sub-Saharan Africa.
Implications for Policy and Practice
Ensuring sustainable strategies for equitable teacher allocation requires both a policy environment that prioritises data-driven decision-making and a culture that fosters innovation and learning among policymakers. This study demonstrates how collaborative work with government can enable research to have a major influence on policy, supporting a cost-effective response to the teacher shortage crisis.
Outputs Register
The outputs register lists all of the products produced as part of the EdTech Hub-commissioned research project on the impact of GIS-supported teacher allocation in Sierra Leone. This collaborative effort involved the contributions of EdTech Hub, Fab Inc, and the Education Workforce Initiative.
Study Team
- Björn Hassler, Principal Investigator
- Taskeen Adam, Co-Principal Investigator
- Paul Atherton, Co-Principal Investigator
- Katie Godwin, Co-Principal Investigator
- Madleen Frazer, Lead researcher
Previous contributors to this study:
- Chris McBurnie, Lead researcher
Key Partners
