Play to Learn - Improving Foundational Learning with Technology Aided Formative Teaching

• In Event: Unlocking the potential of Quality EdTech to support Learning at Home

Proposal

The role of educational technology (ed-tech) or computer-aided learning (CAL) in bridging educational gaps has garnered substantial interest from researchers and policymakers (see Escueta et al. 2017 for a review). CAL has the potential to alter learning outcomes by expanding access to education, augmenting traditional classroom training, delivering personalized materials, and encouraging engagement through compelling content. Some studies have shown that ed-tech improves students' educational achievements significantly by providing adaptive learning opportunities (Muralidharan et al. 2019; Angrist et al. 2022). However, the efficacy of ed-tech has not been tested on younger children during their most formative years. In addition, the delivery model has either been phone and SMS or an app used in centers. Teacher-assisted remedial use of ed-tech where formative assessment is central has not been tested. Summative assessment learning models that focus on end-of-year tests require less teacher effort but do not give teachers the opportunity to offer remedial assistance to struggling students. On the other hand, formative assessment empowers teachers to focus on marginal children. However, the cost of effort for the teacher to determine each child's learning needs and how they are changing can be very high. Ed-tech products have the potential to reduce the cost of effort for formative assessment for teachers. Yet, little is known about the impact of such products on learning outcomes.

This study examines the effect of a teacher-assisted ed-tech app’s integration in schoolwork that facilitates a formative learning model on first and second-grade students' test scores in India. We leverage the deployment of an educational app named Chimple. The app is available in the local vernacular (Hindi) and focuses on foundational learning (literacy and numeracy) with a game-based approach to boost children's engagement. Additionally, the Chimple dashboard tracks students' engagement levels and progress to aid formative assessment, which is diagnostic in nature, rather than summative assessment, which is evaluative.

To assess the impact of Chimple on children's test scores, we use a stratified randomized control trial where a school grade is the unit of randomization. Our sample consists of students in grades 1 and 2 in 34 schools managed by the Bharti Foundation, across 5 districts in Haryana, India. We used average test scores in school assessments to group schools in 8 strata. In each stratum, 50% of schools were randomly chosen to offer the Chimple program in grade 1; the other schools offered the Chimple program in grade 2. The grades that did not receive Chimple in the school served as the control group.

Normalized to the baseline control mean, we find a 0.25 standard deviation improvement in math scores on an independently-administered test. This is driven by a 0.33 standard deviation improvement in grade 1. For children with scores in the lowest quartiles of the baseline math test, we find an effect of 0.45 standard deviations. English scores for grade 2 students improved by 0.23 of a standard deviation. Our results have significant policy implications for designing low-cost adaptive remedial education in developing countries.

Author

• Sheetal Sekhri, University of Virginia