Personalized Education at Scale: Evidence from a Randomized Field Experiment in India
38 Pages Posted: 30 Mar 2016 Last revised: 14 Oct 2022
Date Written: August 10, 2018
Abstract
Personalized learning using computers requires students to access digital devices such as laptops or tablets to play videos and execute software applications. Typically, students in developing countries do not have access to such devices. Therefore, implementing personalized education at scale is a challenge. We establish that the personalization of homework questions can improve learning in this context. Students can be served personalized questions based on their performance in previous questions on mobile phones widely available in developing countries, thus enabling homework personalization at scale. We randomly assigned personalized homework to half the students and traditional homework to the rest in two schools in India. We found that in standardized exams: (1) Students in the treatment group obtained 0.16σ higher scores; (2) higher personalization led to higher scores in exams; and (3) students with medium abilities gained the most. We contribute to the literature on the value of Computer-Assisted Learning in improving education delivery under scarce resources.
Keywords: Economics of Education, Education Production Function, Personalized Learning, Adaptive Learning, ICT and Education
Suggested Citation: Suggested Citation