Putting the horse before the cart: Why materials are foundational to effective education systems

• In Event: What Makes Literacy Work? Materials, Access, and Practice

Proposal

This paper estimates the causal effects of an intervention designed to address deep inequities and uneven distribution of educational resources in Northwest Nigeria. In this region, fewer than 28% of children and youth achieve basic learning levels, with progress hindered by limited teacher training, subpar instructional methods, scarce educational materials, overcrowded schools, poverty, and low parental education (Humphreys, 2015; National Population Commission, 2016). Access to textbooks is especially unequal: less than half of children have English or math textbooks, and only 13% have local language textbooks (RTI International, 2016). To help reduce these disparities, the UK-funded Partnership for Learning for All in Nigeria (PLANE) is implementing a seven-year (2021–2028) early literacy and numeracy program across 1,953 schools in three northern states. PLANE’s approach combines teacher training with provision of literacy and math materials to improve student outcomes in some of the most resource-constrained settings.

This study explores whether learning materials enhance student learning outcomes. Existing research finds that provision of learning materials does not guarantee learning gains, in part because of language mismatch or lack of proper use at the school level (Glewwe & Muralidharan, 2016; Murnane & Ganimian, 2014). Thus, our paper adopts a quasi-experimental approach to explore the direct effects of providing local language materials to both students and teachers, coupled with teacher training on their use. In addition, we explore equity implications by testing whether the intervention impacts vary by sex or socioeconomic standing.

The study takes place in three Northern states in Nigeria, Jigawa, Kano, and Kaduna, spanning 17 Local Government Areas (LGAs), which are the equivalent of district-level subnational units. The study includes second grade children in formal primary schools, the majority of whom speak Hausa as their first language. We measure reading and math performance using the Early Grade Reading and Math Assessment (EGR/MA) developed by RTI International, which has been used globally and has been adapted to the Hausa language (Dubeck & Gove, 2015). The analytic sample includes a total of 224 schools across 47 LGAs, 111 of whom are PLANE implementing schools and 113 are control schools from neighboring LGAs. We sampled approximately 6 second grade students at each school. Additionally, we collected student assessment data across 4 rounds spanning the beginning and end of the 2022-23 and 2023-24 school years, respectively.

PLANE districts were non-randomly selected by program staff and state education officials; as a result, a randomized trial was not possible. We therefore use a spatial regression discontinuity (SRD) design as the empirical strategy to estimate impacts of the intervention. The SRD relies on geographic proximity of schools to district borders as the basis for assignment. In this case, the LGA borders represent sharp cutoffs for the implementation of PLANE. This methodology relies on the condition that schools and students are similar on either side of the border, thus mimicking the properties of a randomized trial. We supplement the analysis by incorporating a pre-intervention baseline and multiple post-intervention periods, enabling us to trace the evolution of the program impacts over time (Butts, 2023).

We find that the provision of student and teacher learning materials, coupled with teacher training on material use, has significant positive impacts on students’ reading and math outcomes. Because of the data available across multiple time periods, we provide some insight into the temporal dynamics associated with the intervention’s impacts. We find that the intervention showed no significant impacts in the first year of implementation, but these impacts then become positive and significant in the second year. This is true for both reading and math outcomes. In the first year, we estimate reading and impacts that range between -.18 and .13 SD, depending on the subgroup and assessment subject, while in the second year, the estimated impacts range between .24 and .63 SD, respectively.

Overall, our results demonstrate that provision of learning materials can be effective in yielding significant improvements in student learning in low-resource contexts. Our findings do not dismiss earlier research results but suggest that while learning materials alone may have limited impact, their effectiveness is significantly enhanced when combined with appropriate contextualization and teacher training, thereby maximizing the potential benefits of similar educational interventions. The paper also provides insight into how teacher/institutional experience with the intervention plays a role in realizing the potential positive effects of the intervention. This finding reflects the need for time for teachers to acclimate and adjust their teaching routines and lesson plans to newly provided materials.

Authors

• Wael Moussa, FHI 360

• Rachel M Hatch, FHI360

• Caitlin Kim

• Elizabeth Adelman, FHI 360

• Emily J Koester, FHI 360