Jing Liu

Current Position: 

Post Doctoral Research Associate at Brown University's Annenberg Institute

Jing Liu was a Ph.D. candidate in Economics of Education at Stanford, where he also earned a M.A. in Economics in 2016. He earned his B.A. in Economics in 2011 and M.A. in Economics of Education in 2013, both from Peking University, China. His research interests focus on using computational social science methods, especially text-as-data methods, and high resolution data to understand the mechanism of education production and inform education policy making. He has published articles in Journal of Policy Analysis and Management and AERA Open.

His dissertation, which won the 2017 National Academy of Education/Spencer Dissertation Fellowship, examines the roles teachers and peers play in education production. The first Paper examines how peer interaction affects student learning outcomes and persistence in online courses offered by a large for-profit university by using detailed written communications between students in online discussion board. The second paper measures teachers’ contribution to student engagement in secondary school using students’ unexcused class absences as a proxy. After creating value-added to student attendance, this paper further investigates how this new dimension of teacher effectiveness influences student high school graduation and dropout above and beyond teachers’ impact on student test scores. The third paper uses “text-as-data” methods to create metrics of teacher practices in English language arts classrooms at upper elementary school. Different from the conventional approach of teacher observation, this study quantify teacher behaviors through their language and interaction with students from word-to-word transcriptions of classroom videos. These teacher practices are then compared to teachers’ value-added scores and classroom observation scores. These three papers together exemplify how the use of high-resolution data from education processes, rigorous causal methods, and cutting-edge computational social science tools can significantly advance our understanding of education production.