Working Paper: NBER ID: w15867
Authors: Eric A. Hanushek; Ludger Woessmann
Abstract: Critics of international student comparisons argue that results may be influenced by differences in the extent to which countries adequately sample their entire student populations. In this research note, we show that larger exclusion and non-response rates are related to better country average scores on international tests, as are larger enrollment rates for the relevant age group. However, accounting for sample selectivity does not alter existing research findings that tested academic achievement can account for a majority of international differences in economic growth and that institutional features of school systems have important effects on international differences in student achievement.
Keywords: international student achievement; sample selection; economic growth; educational production functions
JEL Codes: C83; H4; I20; O40
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| sample selectivity (C34) | test scores (C52) |
| test scores (C52) | economic growth (O49) |
| sample selectivity (C34) | economic growth analyses (O57) |
| sample selectivity (C34) | test scores and economic growth (O49) |