Literacy Boost Senegal Baseline Report

This report examines the results of a baseline student background survey and reading assessment conducted in November 2014 as part of Save the Children’s Literacy Boost program in partnership with World Vision Senegal. The survey and reading assessment included 817 grade 2 learners throughout 42 schools in three Area Development Programs (ADPs) covered by World Vision Senegal. The 42 schools are split into 21 primary schools designated to receive Literacy Boost in year 1 of the program and 21 Comparison primary schools which will receive no intervention until a later stage. School headmasters were surveyed to assess the level of resources available at each school and to get an idea of community and parental perceptions of literacy in these three ADPs. There were three primary purposes of this assessment. The first was to examine the current French literacy skills of children in the target ADPs. The results show that grade 2 students have low mastery of foundational French reading skills—over 75% of students could not identify single letter, and almost all (99%) of students could not read a grade-level reading passage. Literacy Boost programming in the school and community must start out with an emphasis on foundational reading skills.

The second purpose of the assessment was to determine whether the Literacy Boost and Comparison groups are statistically similar in order for the impact evaluation to be successful. In terms of background characteristics, there are only two statistically significant differences between Literacy Boost students and Comparison students: students in schools that will receive Literacy Boost are more likely to have repeated grade 1 and are more likely to have seen someone at home reading in the last week. In terms of baseline reading abilities, students in schools that will be part of Literacy Boost significantly out-scored students in Comparison schools on letter identification, most-used word identification, invented word recognition, and listening comprehension. Differences such as these could be cause for concern about the authenticity of the random selection process and whether students in the two groups are similar enough that the Comparison group is a valid counterfactual for the evaluation. However, we have reason to believe that the assumptions for our identification strategy still hold and that the Comparison group still serves as a valid counterfactual for the impact evaluation. A more

detailed explanation of this is provided in Section 6.7. Finally, the data were also used to explore whether students’ reading skills vary by certain dimensions of equity, including sex, preschool attendance, socioeconomic status, home literacy environment, and chore load. Of the equity measures that were tested, only home literacy environment is suggestive as a factor that may influence whether different groups of students learn equitably.