Randomized Impact Evaluation of Various Early Reading Skills Interventions in Mozambique

October 3, 2016

Improving teaching is a big challenge around the world. Some studies have shown that when teachers and school administrators have clear information about student performance, they are able to make better decisions for improving learning. For example, grouping students by performance level or adopting different teaching strategies. The Government of Mozambique implemented a pilot program to raise student achievement in reading through new pupil testing requirements. Researchers evaluated the effectiveness of this strategy for improving reading outcomes in young children.


Research area: Education
Country: Mozambique
Evaluation Sample: 110 primary schools
Timeline: 2012 - 2015 (Completed, endline report pending) 
Intervention: Assessment tests, training, information 
Researchers: Sophie Naudeau, World Bank; Marie-Helene Cloutier, World Bank; Maria-Jose Ramirez, World Bank
Partners: Government of MozambiqueMinistry of Education, MozambiqueDepartment for International DevelopmentAmerican Institutes for Research


Policy Issue

Early reading forms the foundation for future learning. Children who fail to develop basic reading skills early in life are likely to lag further as instruction gets more complex. While experts and governments understand the importance of breaking the cycle of low student performance early, rigorous evidence on cost-efficient ways to improve early reading for a large population of children remains scarce. 


Mozambique, a southeast African country emerging from a period of civil war and natural calamity, has made tremendous progress over the past decade in opening access to primary education. But scores on standardized reading tests are very low. Portuguese is the official language of Mozambique and the language being tested, but many students do not speak Portuguese at home.

Now that enrollment is strong, the country has shifted its emphasis to focus on improving the quality of primary education. This evaluation looked at classroom-based assessment as a route for giving teachers, students and their families information for improving reading skills in Mozambique.

Photo: © Curt Carnemark / World Bank

Intervention and Evaluation Details


This impact evaluation looked at one main intervention to improve reading performance of young children and seeks to provide evidence-based information on its impact to implement a program on a larger scale. A tailored version of the Provinha reading assessment program, developed in Brazil, was at the core of this intervention. The twice-yearly Provinha test comprises 20 multiple-choice questions given to 3rd graders to assess reading skills they should have mastered after two years of primary school. In addition, the test provides teacher with a guide to the reading skills their students should be learning.

Starting in 2013, the Provinha assessment was introduced to increase teacher awareness of their students’ strengths and weaknesses in reading, to familiarize teachers with the new practice and to improve student outcomes.


The evaluation design was experimental using a randomized controlled trial. The study focused on one district in the each of two Northern provinces, Tete and Niassa. Researchers randomly assigned 52 schools to a treatment group and 58 to the control group. (The rest of the schools in these districts participated in Provinha testing, but were not part of the evaluation.) In the treatment group, teachers administered and scored the Provinha standardized reading test and the school directors participated in a workshop to analyze their school’s results and discuss strategies for improvement. In the control group, an external firm administered and scored the Provinha test without involvement by teacher or school officials.

Baseline tests were administered in the first two months of the school year, with follow-up tests about six months later. The primary outcomes analyzed are the reading skills scores for 3rd graders taking the Provinha test. Researchers also surveyed school directors, teachers, parents, and community-leaders and observed classroom instruction to collect information on teacher and director assessments of student reading skills, teacher motivation and qualifications, parental involvement and support, and student and teacher absenteeism.

Final results are pending.