This file contains data on test scores in reading (Leblango and English), oral English and math from students participating in NULP from 2013 to 2017. The dataset consists of 73,441 student-year observations. Variable description: PupiID: Pupil identifier; allows for identifying students across years. SchoolID: School identifier. Group: Stratification group (School Level) Cc: Coordinating center School_Type: (1) School sampled in 2013 (38 schools); (2) School sampled in 2014 (90 schools); (3) Pure control schools (30 schools). Sample_Type: (1) Student sampled beginning of the year (baseline); (2) Student sampled end of the year (endline). Year: Year when tested Wave: (1) Tested in the beginning if the year (baseline); (2) Tested in the end of the year (endline) Grade_Level: Grade level when tested. Cohort: (1) student sampled in 2013; (2) student sampled in 2014; (3) student sampled in 2015; (4) student sampled in 2016. Stream: Stream Age: Student age. Gender: Student gender Leblango_EGRA: Early Grade Reading Assessment (Leblango) English_EGRA: Early Grade Reading Assessment (English). Oral_English: Oral English scores Math: Math scoresLiteracy is the foundation for an informed, skilled citizenry. But in East Africa, less than 1/3 of pupils possess basic literacy skills. Ugandan children perform the worst; only 44.5 percent pass basic literacy tests. Early Grade Reading Assessment (EGRA) data from a Research Triangle Institute survey in northern Uganda in 2009 indicated that 82% of P2 pupils could not read a single word in the local language, compared to 51% of P2 pupils in the central region. Similar to other African countries there are many problems in Uganda's education system, including undertrained teachers, lack of materials and quality methods for teaching literacy, non-existent systems for tracking pupil performance, and parents, communities and local officials that lack the know-how to support and advocate for their children's education. Despite strong mother tongue education policies, due to underdeveloped orthographies and a lack of materials in many languages, implementing successful mother tongue literacy programs poses a significant challenge for African countries, including Uganda. While many educational interventions and literacy programs have been implemented in Africa, impacts have been minimal overall; moving to scale has also proven problematic as program effects reduce further. Since 2010 Mango Tree, a private, locally owned educational tools company, has been piloting a successful early literacy project in one language community in northern Uganda. The main goals of the Mango Tree program include increasing literacy rates, enhancing education quality through improved, effective materials and teachers, and fostering a culture of reading among pupils, parents and communities within a cost-effective and scalable framework. Compelling evidence for the large benefits and cost-effectiveness of the intervention comes from a pilot randomized evaluation of the program conducted by University of Michigan researchers in 2013 and 2014. The Literacy Laboratory Project (LLP) will scale up and evaluate the Mango Tree literacy program, whose model delivers better-quality teacher instruction, access to relevant literacy materials, inclusive approaches to learner assessment, parental and community engagement in schools and strengthening literacy infrastructure so that reading and writing, especially in local languages, becomes a meaningful part of daily life in households and communities. This scale-up will test a piloted and improved model to evaluate its effectiveness and test the mode of program delivery. Under the LLP, researchers from the University of Michigan will conduct a rigorous randomized control trial of the program in the Lango Sub-region over 4 years to measure the effectiveness of the instructional model, teacher training and support supervision innovations and literacy materials and methods on Primary 1- Primary 3 pupils' literacy achievement and explore public-private avenues for scale-up. We will study 128 schools, which are randomly assigned to either the full LLP implemented by Mango Tree's field officers, a partial-program implemented by Government Teacher Tutors, or a control group. The study will also randomize instructional materials to evaluate their contribution to effective teaching. The study will collect a rich set of pupil, parent, teacher, classroom, and school-level longitudinal data. Learning will be measured principally in terms of improvements in EGRA and Early Grade Writing Assessment scores. Our goals are to: 1) demonstrate that big effects on learning are possible (as the 2013 pilot evaluation results point toward); 2) show that with the right combination of training, teaching and learning materials and correct support, teachers can be supported to effectively teach literacy - even in rural, under-resourced, overcrowded classrooms; and 3) to test and evaluate economic approaches to implementation at scale to determine value-for-money impacts on pupil learning and teacher performance in African schools.

Early grade reading assessments using student surveys. Students were interviewed face to face in schools A sample of schools were sampled in 2013 (38 Schools) or in 2014 (an additional 90 Schools for a total of 128). Students were then randomly sampled either in the beginning of the school year or end of school year. Students were sampled in first grade.