SREE 2017 Spring Conference
March 1-4, Washington, DC
"Beyond the Average Treatment Effect: Differential Effects of an Intervention on Growth Trajectories of Literacy Skills as a Function of Student Characteristics"
Authors: Young-Suk Kim & Benjamin Piper
The primary goal of this study is to examine whether the effect of a comprehensive literacy intervention varies longitudinally as a function of student characteristics such as gender, socio-economic backgrounds (SES), exposure to early childhood education, and maternal literacy. Of particular interest was the relations of these student characteristics to variations in the growth rates in literacy skills. The children in the sample were drawn from three of Kenya’s counties as well as low cost private schools in Nairobi’s so called informal settlements. A total of 628 children (312 girls; 357 in treatment condition) participated in a comprehensive literacy and numeracy intervention in Grades 1 and 2. Clusters of schools were randomly selected and then randomly assigned to treatment and control groups. At the school level, simple random sampling was employed so that the children selected were representative of the entire Grade 1 population not simply one class at the school. Children’s literacy skills were assessed at three time points: before, middle and end of the intervention, which corresponded to the beginning of Grade 1, end of Grade 1, and end of Grade 2. Different patterns were found for students in treatment versus control conditions. For students in the treatment condition, SES and early childhood education were consistently and positively related to growth rates in the literacy skills. For students in the control condition, none of the student characteristics were related to individual differences in growth rates in literacy skills. The present results indicate that the effect of a comprehensive literacy intervention for primary grade children in Kenya was not uniformly distributed across students. Instead, certain students benefitted more than others – those from higher SES and those who received early childhood education learned more from systematic and explicit literacy instruction, after accounting for gender and maternal literacy. On the contrary, for students who were not exposed to systematic and explicit instruction in literacy, students’ background characteristics did not predict growth rates.