World Science Day for Peace and Development highlights the important role of science in addressing society’s needs and promoting innovation. In the education sector, scientific research leads to technological developments that can enhance teaching and learning in the classroom. With the advent of artificial intelligence, the role of this and other emerging technologies in learning increasingly dominate discussions on education innovation. At the Luminos Fund, however, we believe that developing new tools and processes for measuring learning in real-time is a vital way for science to improve education – and is essential to ensure all children can succeed and thrive.
Below we share reflections from our Data Scientist, Dr. William Chantereau, on the critical role of data and science in driving innovations and learning at Luminos.
What do you believe is the role of science in driving innovations in education?
Two fields have significantly shaped my journey into the education sector: astrophysics, where I began my career, and later, data science. Through these experiences, I have come to strongly believe in the critical role of data in ensuring a lasting impact in the education sector and beyond. As a foundational pillar of science, data provides the empirical evidence necessary to inform decision-making, drive innovation, and solve today’s most pressing global education challenges.
Dr. William Chantereau, Luminos Data Scientist
What do you see as gaps in the sector around data collection and innovation?
Recently, I read an article by the Center for Global Development that revealed significant gaps in understanding among policymakers about the scale of the learning crisis. Too often, there is a disconnect between perception and reality – which is why data is essential. In my role, I help Luminos use data as a powerful tool for creating immediate and tangible improvements in children’s learning. Luminos collects several types of data, including real-time classroom observation data, evaluative math and reading assessment data, demographic data, and external evaluation data. This helps us identify what is and isn’t working in the classroom and make the necessary adjustments.
A key challenge in the sector is collecting trustworthy and meaningful data to ensure impact, particularly in the Global South. The fact that less than 50% of African countries have even the basics of education data on out-of-school children is proof that there are significant gaps in the full range of types of data necessary to ensure learning for all children. Such gaps in education data greatly undermine efforts to achieve equity within countries. For example, only one in three studies of co-educational interventions disaggregate impact by gender. Without clear data on specific learning trends by gender, how can we adequately support both girls and boys to thrive in the classroom and beyond?
How does Luminos build the capacity of community partners to collect and input data?
Luminos has invested in a dedicated team of data specialists and country-level data leads, known as the Impact team, solidifying our role as a data-driven organization in a field that often lacks such focus. A key objective for the Impact team is fostering a similar data-driven culture with our community-based partners. We provide continuous coaching and support, holding regular meetings to review and discuss partners’ data with a focus on accuracy and consistency.
Most importantly, we are always innovating our approach. We have invested in digital data collection and analysis tools that can scale and do not rely on complicated tech processes, making it easy to incorporate changes. For example, our data dashboard tool allows us to quickly input data for thousands of students each year, in each of our countries, building on existing data sets. These tools are durable enough to handle increased volumes of data, as our programs grow. We support partners to use these dashboards to monitor daily progress across classrooms and make adjustments when necessary.
As our community partners build capacity over time, we have observed significant progress in data quality, reflecting their growing proficiency and commitment.
Supervisor Kennedy congratulates a student in Ghana after completing a weekly numeracy and literacy assessment. Kennedy has received training on how to correctly administer the assessment and enter student scores in Kobo. Such trainings help establish the data’s validity and quality. (Photo by Obeng Baah for the Luminos Fund)
How does Luminos use data to drive impact?
Collecting high-quality data is just the first step – its true value is realized when we use it effectively. A lot of my day is spent combing through data and asking questions. What exactly are we looking at? What does the data tell us about learning? One of the key insights we often glean from our data is pinpointing specific skills that students struggle with, such as blending letter sounds or recognizing and working with double-digit numbers. This enables us to prioritize and focus on these areas to ensure timely support for students’ learning needs. At the classroom level, high-quality data enables teachers to quickly identify struggling learners and make necessary adjustments to ensure that every child is learning effectively.
Each year, we analyze data and leverage the insights to refine our curriculum, coaching, and support to community partners and teachers. This ongoing commitment to iteration drives the positive outcomes of our programs year after year. Beyond our own programs, this commitment to building a robust evidence base helps build education systems that truly address the needs of each learner.
Learn more about Luminos’ approach to using data to drive impact in the Real-Time Data element of the Luminos Method!
We have substantial new data from Luminos program countries this year. What, if anything, about these results stands out to you? What are your key reflections?
Each year, we commission external evaluations of learning in our country programs using the Early Grade Reading Assessment (EGRA) and Early Grade Mathematics Assessment (EGMA). This measures key literacy and numeracy skills, including the number of correct words per minute read by students. In the 2023-24 school year, Luminos achieved its highest-ever impact across all four programs in Sub-Saharan Africa, far exceeding national averages on correct words per minute.
I remember when our programs just started out in The Gambia and Ghana, we had some initial challenges with low learning achievements. What’s really impressive about our focus on data is how we have been able to use learnings, including from our more established programs in Ethiopia and Liberia, to improve our impact in these new programs in a short amount of time. In Ghana and The Gambia, students have nearly doubled their scores in just one year – an unparalleled and truly remarkable achievement.
Increasingly, across our programs, we have the data we need to understand what works and what doesn’t, allowing us to make informed improvements. As we accumulate larger datasets on learning outcomes, demographics, attendance, and more, we increase our ability to detect important patterns, such as correlations between learning outcomes and demographic data. This enables us to provide timely and tailored support to students at the early stages of the program.
Learn more:
- Discover Luminos’ approach to collecting, analyazing, and disseminating real-time data in the Luminos Method element>
- Explore the latest EGRA/EGMA results from our government adoption with structured pedagogy program in Ethiopia here>
- Read a new peer-reviewed journal article highlighting RCT results from the Luminos program in Liberia here>
Dr. William Chantereau is a Data Scientist for the Luminos Fund where he analyzes program data, builds tools to automate data processing, ensures data integrity, and oversees critical data management processes across the organization. Under William’s leadership, Luminos was selected to participate in the 2023 LEAP project, a global initiative of the Jacobs Foundation and MIT Solve aimed at advancing evidence-based solutions to help children reach their full potential.
William previously worked as a data scientist at Vendigital in London, a top-tier management consultancy working to deliver data-led accelerated cost transformation. Before that, William worked as an astrophysicist in France and the UK. William holds a PhD in Astrophysics from the University of Geneva (Switzerland) and a master’s degree in Fundamental Physics from the University of Toulouse (France).
