Improving Literacy Instruction in Kenyan Classrooms: Results From the Health and Literacy Intervention (HALI) Project

Literacy levels for primary school children across sub-Saharan Africa, including Kenya, remain low. While many issues such as poor health, poverty, and limited access to educational materials likely contribute to delayed reading acquisition, Kenya’s education policy does not mandate a specific method of literacy instruction and only states that instruction should meet students’ learning needs and lesson objectives (Ministry of Education 2006). Furthermore, limited resources constrain training opportunities for teachers. As a result, there is likely variability in the quality of literacy instruction school children receive.

This study assesses the impact of a teacher training intervention to improve first grade teachers’ literacy instruction in coastal Kenya government schools. The intervention included a three-day training, a manual with 140 lessons, and weekly support through text messaging. Participants were 98 first grade teachers (50 randomized to the program and 48 to the control group). Teachers’ were assessed through direct observations by trained assessors midway through the year, using an adapted version of the CLASSIC observation tool to assess pedagogy for language instruction (Scanlon et al., 2003). Teachers were rated across four dimensions of instruction including materials used, instructional focus, and teacher and student behaviors.

Observations of control group teachers reveal that teachers spent a small amount of time teaching phonics(i.e., stressing letter-sound relationships) during literacy instruction. On average, teachers spent less than 5% of class time teaching phonological elements of words (e.g., letters, sounds) and only 14% teaching skills of putting letters, sounds or syllables together (blending) and taking them apart (segmenting) to read or spell words. 

The intervention increased the amount of time teachers spent using written materials (b=0.45, p<.01) and reduced oral instruction (b=­-0.24, p<.05). Teachers increased time teaching letters and sounds (b=0.38 and 0.25, p<.05) and blending and segmenting of words (b=0.32, p<.01). Student participation also increased; students spent more class time reading (b=0.36, p<.01) and less time disengaged in classroom activities (b=-0.19, p<.10). Further analyses will group teachers based on instruction across all categories using cluster analysis and assess intervention impacts on patterns of literacy instruction. In addition, analyses will assess if changes to literacy instruction were related to students’ literacy outcomes.

This study provides empirical data about literacy instruction in Kenyan classrooms and how to improve teaching quality. Understanding how to change instruction to improve students’ literacy outcomes is critical to ensuring quality education for all students.