Abstract

Olaf Köller

Comparing Apples with Oranges? Linking National and International Large-scale Assessments

National assessments have increasingly been linked with international assessments in order to gain insight into the different results of the studies or to use the international benchmarks in national studies. The aim of this research is to classify national assessment studies in Germany according to international scales as used in the Trends in International Mathematics and Science Study (TIMSS) and the Programme for International Student Assessment (PISA). As an example mathematics achievement data from the National Educational Panel Study (NEPS) in Germany is used to link it to the criterion-based frame of TIMSS. One requirement for a strong linking is a high conceptual overlap of the frameworks in mathematics. Therefore, the frameworks and test specifications of the two studies are compared. The results indicate a high degree of overlap between the two studies. To link the studies statistically, two different methods – the equipercentile equating and an IRT linking approach – are compared regarding their classification accuracy according to TIMSS international benchmarks (proficiency levels). The result of the linking is that both methods show similar descriptive statistics and satisfying classification consistency. Nevertheless, the equipercentile equating has small advantages and is therefore more suitable. Inferences are possible on a population level, but the results should not be reported or interpreted on an individual student level.

Published Feb. 25, 2015 12:06 PM - Last modified Feb. 25, 2015 2:03 PM