English title: Confusing Credentials: The Cross-Nationally Comparable Measurement of Educational Attainment
Author(s): Silke L. Schneider -
Type: Thesis / dissertation
The quality of educational attainment measures lies at the heart of many cross-national micro-sociological research projects and international education statistics. This study aims at validating cross-nationally comparable measures of educational attainment, among which are the International Standard Classification of Education 1997 (ISCED 97) and years of education. Following a conceptual discussion of what educational attainment means, the most common ways of measuring educational attainment cross-nationally as well as previous evaluations thereof are reviewed. Then, the implementation of ISCED 97 in cross-national surveys is examined by looking at both the resulting educational attainment distributions in three European surveys as well as the data generation and harmonisation process. Finally, a number of cross-national measures of educational attainment are compared with country-specific measures with respect to their information content by firstly examining the dispersion of educational attainment, and secondly the predictive power when explaining the core social stratification outcomes, occupational status and social class attainment, by educational attainment. The main results of the study are that the measurement of educational attainment in cross-national surveys is affected by a number of avoidable weaknesses which adversely affect the validity of claims based on analyses of these data: 1. Countries and surveys are inconsistent in the way they measure educational attainment and apply ISCED 97 to national data; and 2. actual years of education and the one-digit version of ISCED 97 distort measures of association to differing degrees in different countries, making cross-national comparisons using these measures highly problematic. Therefore, some amendments to the implementation of ISCED 97 in cross-national surveys and coding for statistical analyses are proposed. As part of the latter, an alternative simplification of ISCED 97, optimised for European survey research, is developed and validated. Moreover, suggestions for data collection procedures are made to improve the measurement of educational attainment nationally and cross-nationally.
Awarding institution: University of Oxford
Number of pages: 485