Article ID Journal Published Year Pages File Type
5055450 Economic Modelling 2011 8 Pages PDF
Abstract
► International school comparisons statistically test for "significant" differences. ► These tests usually do not take the multiplicity of tests performed into account. ► This approach may overstate the differences between countries' schools. ► We use multiple testing techniques to control for this problem. ► Using PIRLS data, we find up to 30% fewer significantly different countries.
Related Topics
Social Sciences and Humanities Economics, Econometrics and Finance Economics and Econometrics
Authors
,