Instructional coaching and literacy improvement at national scale: Lessons from Kenya’s Tusome early grade reading activity

• In Event: Highlighted session: Coaching to improve early grade reading instruction: evidence on effectiveness and sustainability

Proposal

The Tusome Early Grade Reading Activity is USAID’s flagship education program in Kenya. Operating fully at national scale, the program has driven significant reading improvement in English and Kiswahili for over 6 million grade 1-3 children in more than 23,000 public primary schools and 1,500 super-low-cost private schools in Kenya’s slum areas. The program has realized these outcomes through a combination of book provision (more than 22 million books), teacher training (delivered to more than 100,000 teachers and head teachers multiple times per year), deployment of ICT (including the open-source Tangerine:Tutor platform), and instructional coaching.

In academic 2017 the Curriculum Support Officers (CSOs) conducted observations of nearly 130,000 reading lessons. Of those, more than 105,000 observations met Tusome’s criteria for a coaching visit:
1. They took place in Tusome schools, classes, and subjects;
2. The observation forms were fully completed;
3. The observation duration was between 25-100 minutes;
4. The lesson took place on a weekday between 8:00 and 15:10 hours;
5. GPS coordinates were present and matched the known location of the school; and
6. The record included brief assessments of 3 pupils’ reading skills.

This presentation shares the findings from research currently in progress to analyze Tusome’s 2017 lesson observation data. The research aims to answer the following research questions, centered on three key elements of the coaching interaction:
1. Do indices of effective teaching practices (both literacy-specific and related to general pedagogy) show evidence that instructional delivery improves over the course of an academic year after repeated coaching interactions?
2. Is additional experience with the instructional methodology associated with better instructional delivery, as measured by indices of effective teaching practices?
3. Does the use of moderately scripted lesson plans eliminate variability in instructional pacing?

In the context of the panel discussion, the authors will highlight lessons that can be drawn from Tusome’s experience by implementers of other large- or national-scale instructional interventions. These will include:
1. Designing a coaching program to operate at scale;
2. Effectively combining incentives and sanctions to drive coaching activities; and
3. Effectively combining automated, moderately high-tech data pipelines with qualitative, low-tech feedback on coaching of coaches.

Submitted under the Applied Research rubric, the presentation is highly relevant to current and future implementers of large-scale reading improvement programs whose theory of change includes improved instructional delivery. It explicitly addresses the development, management, and sustainability of an instructional coaching infrastructure that depends wholly upon government employees and operates within the existing framework of the Kenyan education system. The presentation provides insight into a core function of one of the most successful national-scale reading interventions to date, affording session attendees opportunities to learn from and potentially improve upon Tusome’s success.

Authors

• Timothy Slade, RTI International

• Kim Darnell, University of California, Berkeley School of Information

• Lucy M. Wambari, RTI International

• Alex Dauenhauer, University of California, Berkeley School of Information