Using Assessment for Learning to Enhance Numeracy Instruction- Nigeria, South Africa, and Kenya

Introduction

Launched in 2024, the Daara Development Academy is a co-created initiative designed to strengthen the capacity of African organizations to scale evidence-based solutions for foundational learning. An important component of this program was an Innovation Fund, intended to support collaborative projects between Daara partners that addresses sector needs to improve foundational learning outcomes. Projects could include the development of approaches, tools, and resources that align good practice within the Cohort with the science of teaching principles.

The project “Using assessment for learning to enhance numeracy instruction” was awarded funding through this Fund and brought together four African organizations: The consortium was led by TEP Centre which also led the implementation in Nigeria, while Funda Wande led implementation efforts in South Africa and Zizi Afrique, implementation activities in Kenya. eBASE Africa was responsible for evaluation and evidence generation.

Case for follow-up and scale

The teacher-centered numeracy intervention rooted in error analysis and formative assessment has suggested strong feasibility, clear instructional value, and early signs of impact across three diverse education systems in Nigeria, South Africa, and Kenya. Teachers not only adopted new practices but sustained them. Learner outcomes improved significantly, classrooms became safer, more reflective spaces where learners actively discussed their mistakes, and teachers treated errors as opportunities for learning. With a scalable, low-cost model and a strong foundation of cross-country collaboration, the consortium are poised to expand this approach to reach more schools and teachers across Sub-Saharan Africa and design a rigorous impact evaluation of the program to generate robust evidence needed to embed this practice into national instructional reform.

Overview of the project

This project addresses two critical challenges undermining foundational numeracy in Sub-Saharan Africa: low learner performance in early-grade mathematics and the difficulty teachers face in sustaining improved instructional practices without ongoing support. Evidence from Nigeria, South Africa, and Kenya highlights that, despite training, many teachers struggle to implement effective teaching strategies, and learners continue to face persistent difficulties with key numeracy concepts such as place value and two-digit operations. The proposed solution was to equip teachers with practical tools and structured support to use error analysis as a formative assessment technique. By identifying and interpreting learner errors, teachers can adapt their instruction to address specific misconceptions, transforming mistakes into valuable learning opportunities. The project developed mobile-friendly, evidence-based guides and facilitated collaborative lesson planning tailored to national curricula. Implemented in phases across Nigeria, Kenya, and South Africa, the initiative aimed at building teachers’ capacity, promoting sustained pedagogical change, and ultimately improving numeracy outcomes for early-grade learners in under-resourced settings

Brief overview of the evaluation approach

The evaluation of this proof-of-concept project employed a mixed-methods design to assess the feasibility of the intervention, improvements in teacher capacity, the uptake and maintenance of evidence-based practices, and early signs of student learning. Data collection was conducted at three time points in Nigeria and South Africa, and at baseline and endline in Kenya, focusing on teacher practices, learner engagement, and student performance. Quantitative data were collected through standardized learner assessments, teacher audits, and structured classroom observations, while qualitative insights were obtained from key informant interviews and teacher diaries. Instructional shifts and learner engagement were tracked using audit checklists and observation rubrics, and weekly error logs helped identify common misconceptions to inform lesson planning. The evaluation was further supported by coaching sessions based on audit findings and regular WhatsApp check-ins, offering real-time feedback and support. This triangulated approach provided a robust assessment of the intervention’s feasibility, impact, and emerging evidence of promise in resource-constrained classrooms.

Overview of the findings

Initial results from Nigeria and South Africa, with partial data from Kenya, show clear progress across four key outcomes. First, teacher capacity in error analysis and formative assessment improved significantly. In Nigeria, teachers advanced from basic error detection to diagnosing misconceptions and adjusting instruction. In South Africa, 50% of teachers at endline defined error analysis as identifying, analysing, and understanding learner errors. In Kenya, among the 19 directly trained teachers, 93% corrected learner errors, 86% identified misconceptions, and 100% offered targeted feedback, though 33% struggled to consistently update error logs due to workload. Teachers increasingly moved learners from basic counting strategies to place value methods.

Second, teacher attitudes and behavior improved. In Nigeria, compliance with best practices rose from 36% to 79%; in South Africa, from 16% to 74%, with 71% of teachers reporting greater confidence in using error analysis. In Kenya, 87% facilitated error discussions, 93% adjusted lesson plans based on insights.

Third, student engagement increased. In Nigeria, engagement rose from 49% to 72%, with 75% of teachers observing learners explaining their thinking and 70% reporting peer correction. In South Africa, engagement improved from 56% to 64%, and 62% of teachers reported learners were more willing to explain their reasoning. In Kenya, 50% of learners were discussing mistakes at baseline, 100% felt safe doing so, and peer correction was observed in 50% of lessons at endline.

Finally, there are promising indications that learner performance in foundational numeracy improved. In Nigeria and South Africa, factual and conceptual errors in addition declined, though procedural errors remained, with 78% of procedural errors in Nigeria tied to subtraction with regrouping. In Kenya, where over 80% of errors were factual at baseline, scores rose from 19% to 43%, including 27% to 46% in addition and 12% to 39% in subtraction.

These findings suggest that targeted teacher training in error analysis could potentially lead to more responsive teaching and better learning outcomes, while also highlighting the need for continued support to deepen understanding and sustain progress.

Another notable outcome of the project is the establishment of key collaborations that built pathways for sustainability, notably in Kenya where the project was done in close partnership with the government, providing them with critical insights into learner numeracy challenges that are now being addressed through new approaches to teacher professional development.

Read the full report Here