Returns to effort: experimental evidence from an online language platform

摘要: While distance learning has become widespread, causal estimates regarding returns to effort in technology-assisted environments are scarce due high attrition rates and endogeneity of effort. In this paper, I manipulate by randomly assigning students different numbers lessons a popular online language platform. Using administrative data from the platform instrumental variables strategy, find that completing 9 Duolingo lessons, which corresponds approximately 60 minutes studying, leads 0.057–0.095 standard deviation increase test scores. Comparisons literature back-of-the-envelope calculations suggest can be as effective in-person for college an introductory course.