Individual Submission Summary
Share...

Direct link:

Experiments as Opportunities to Learn: Thinking differently about RCT research on early Grade Reading

Tue, March 7, 11:45am to 1:15pm, Sheraton Atlanta, Floor: 1, Atlanta 5 (North Tower)

Proposal

Deaton (2010) has convincingly argued that RCT research as with other forms of evaluation, is unlikely to lead to scientific progress in the understanding of economic development unless it explores underlying generative mechanisms of change. This paper describes the government-led Early Grade Reading Study which consists of a series of randomised control trials conducted over the past half-decade. The Early Grade Reading Study series has been designed to provide policy insights about the impact of system-wide early grade reading improvement models, particularly the ‘triple-cocktail’, i.e. lesson plans, high quality materials and instructional coaching. Building on the findings of Regression Discontinuity Design/Natural Experiments of Gauteng Primary Language and Mathematics Strategy (Fleisch & Schoer, 2014; Fleisch et al 2016),the programme currently consists of three major RCT studies -- Kwazulu Natal province (system-wide remedial catch-up); North West province (Home Language, Setswana) and Mpumalanga province (English second language). These studies combine base- and end-line learner testing, teacher, principal, and teacher questionnaires with in-depth qualitative school case studies and large-sample lesson observations. The purpose this mix-method research is to address questions of the impact of the basic model overtime, the relative cost-effectiveness of specific components in the model, and the timing and duration of optimal intervention. The multi-site research programme creates the opportunity to the address the common criticism that RCT research is weak on external validity. Across the studies, the core research team and a widening network of researchers are beginning to explore mechanisms associated with teacher enactment and school context.

Authors