Using Random Experiments to Measure the Impact of Computers, the Internet, and Other Forms of Technology on Educational Outcomes

Abstract

This chapter examines empirical approaches to studying the question of whether computers, the Internet, and other computer-related technologies improve educational outcomes. Most studies in the literature examine the question using multivariate regression analysis controlling for detailed school, teacher, student, family, and/or parental characteristics. Because of concerns regarding selection bias a relatively new literature uses randomized field experiments to examine the educational impacts of computer-related technology. Additionally, studies use quasi-experimental approaches such as natural experiments and regression discontinuity designs to estimate educational impacts. This chapter reviews the literature using these approaches and discusses each of the methodologies along with their strengths and weaknesses. Understanding the impacts of technology on education is important, because it sheds light on whether technology is an important input in the educational production process and whether disparities in access to technology translate into educational inequality.