Andy Zhang - RESEARCH

My research focuses on how information, resources, and institutional arrangements shape educational decision-making, instructional practice, and organizational change, with particular attention to inequality and the implementation of policy. In East Asia, I study these questions through families, students, and teachers, examining educational investment, student behavior and learning, and teacher quality. In Latin America, I extend this agenda to schools and education systems, studying how funding, improvement planning, organizational capacity, and accountability shape school-level responses to policy.

East Asia

Student-level: Improving Student Learning Through Behavioral Interventions

This study examines how information interventions can address behavioral barriers to eyeglass use among students in low-income communities. Using a lab-in-the-field experiment involving 8,262 middle school students in rural China, we evaluate four targeted information nudges designed to correct misconceptions and influence students’ preferences for wearing eyeglasses. We find that information about eyeglass prices and the short-term academic benefits of wearing glasses is particularly effective among myopic students who do not own glasses. The effects operate through belief updating and increased awareness, with substantial variation across students based on their characteristics and experiences. Our findings highlight the potential of low-cost, targeted information interventions to improve educational investments and outcomes in resource-constrained settings.

“Information for Vision: Experimental Evidence on Nudging Low-income Students to Wear Eyeglasses,” with Jiusheng Zhu, Yu Zhao, and Xiaoyang Ye. Economics of Education Review. 115 (2026): 102835. [PDF]

Family-level: Family Educational Investment and Decision-Making

This study examines how parents allocate educational resources among their children in response to information about their academic abilities. Using a randomized controlled trial involving 3,750 low-income households with two children across 201 villages in rural China, we provide parents with information about their children's academic performance. By introducing randomized information shocks, the study investigates whether parents adopt complementary investment strategies that favor higher-performing children or compensatory strategies that support lower-performing children. It also examines how investment decisions vary by household socioeconomic status, children's gender, and age. The findings aim to advance our understanding of parental decision-making, human capital investment, and the mechanisms through which family educational investments may contribute to educational inequality.

Teacher-level: Teacher and Instructional Effectiveness

This study examines whether improving classroom instruction through targeted teacher training can enhance student learning in resource-constrained settings. Using a randomized controlled trial involving 68 teachers and 5,120 students across 20 schools in rural China, we evaluate the effectiveness of systematic phonics instruction as an alternative to the traditionally mandated International Phonetic Alphabet (IPA) approach. The intervention significantly improves student achievement by 0.112 standard deviations. These improvements are potentially driven by increased teacher subject knowledge, changes in teaching beliefs and classroom practices, and more positive student learning attitudes. Our findings highlight the importance of evidence-based instructional methods and targeted teacher professional development in addressing learning poverty and improving educational outcomes.

Combating Learning Poverty: Experimental Evidence on Improving Instruction Through Teacher Training, ” with Lu Wang and Xiaoyang Ye. Economics of Education Review. 109 (2025): 102729. [PDF]

Latin America

School-level: School Organizational Behavior and Improvement

This research strand examines how schools make organizational decisions, develop improvement strategies, and translate educational policies into practice. Using large-scale administrative data from Chile, my ongoing research investigates why some planned school improvement strategies are successfully implemented while others are not. By combining natural language processing and machine learning, the study examines how organizational capacity, school characteristics, and the nature of planned activities contribute to differences in implementation. Particular attention is given to how schools respond to institutional demands while operating under varying resource and organizational constraints. This work seeks to advance our understanding of school organizational behavior and identify factors that support effective and sustainable school improvement.

System-level: Education Policy, Governance, and System Improvement

This research strand examines how education systems shape school behavior through accountability policies, performance assessment, and institutional incentives. Using large-scale administrative data from Brazil, my ongoing research investigates how eligibility for standardized school performance assessments influences schools' organizational decisions and resource allocation. Drawing on regression discontinuity designs and machine learning methods, the study explores how schools respond to performance measurement and how these responses vary across institutional contexts. The research also examines whether accountability mechanisms encourage meaningful organizational improvement or primarily changes in measurable school practices. This work aims to inform the design of education accountability systems that promote effective school improvement and more equitable educational outcomes.