Do Value-Added Estimates Add Value? Accounting for Learning Dynamics
Pomona College - Department of Economics
World Bank - Development Economics Research Group (DECRG); World Bank - Development Research Group (DECRG)
Asim Ijaz Khwaja
Harvard University - Harvard Kennedy School (HKS); Center for Research on Pensions and Welfare Policies (CeRP); Bureau for Research and Economic Analysis of Development (BREAD); National Bureau of Economic Research (NBER)
Harvard University - Harvard Kennedy School (HKS)
September 1, 2009
World Bank Policy Research Working Paper No. 5066
Evaluations of educational programs commonly assume that what children learn persists over time. The authors compare learning in Pakistani public and private schools using dynamic panel methods that account for three key empirical challenges to widely used value-added models: imperfect persistence, unobserved student heterogeneity, and measurement error. Their estimates suggest that only a fifth to a half of learning persists between grades and that private schools increase average achievement by 0.25 standard deviations each year. In contrast, estimates from commonly used value-added models significantly understate the impact of private schools' on student achievement and/or overstate persistence. These results have implications for program evaluation and value-added accountability system design.
Number of Pages in PDF File: 44
Keywords: Education For All, Tertiary Education, Secondary Education, Primary Education, Teaching and Learning
Date posted: October 8, 2009
© 2015 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo3 in 1.219 seconds