Sampling for TEDS-M

Abstract

The Teacher Education and Development Study in Mathematics (TEDS-M) is the first large-scale assessment in teacher education using statistical sampling. The sampling design applied in the study allows the computation of correct, precise, and unbiased population estimates for all characteristics investigated in the survey. This chapter introduces non-statisticians to the basic concepts of statistical sampling and its application in TEDS-M using examples and graphical illustrations. Readers will be made familiar with the features of complex samples in general and with the TEDS-M survey in particular. Implications for the conduct of statistical analysis of TEDS-M data and the interpretation of the results will be presented. TEDS-M comprises four target populations, namely, teacher preparation institutions, future primary/lower secondary math teachers, and their educators, for whom reliable estimates of main characteristics were required. The demanding study goals—both combining four target populations into one survey and capturing the complexity and differences of the teacher education systems in the 17 participating countries—posed particular challenges in the design of a multi-purpose international sampling plan. Experiences gained throughout the implementation of this study provide a valuable contribution to the specification of sampling designs for future studies in higher education.