Abstract
This study uses Rasch modelling to link student outcomes over the decade since the introduction of chance and data into the curriculum of an Australian state in 1993. Although improvement is observed over time for intact groups of students, and between grade levels in a given year, improvement across cohorts for given grades over time is not observed. The distribution of the items used in the 2003 survey across the statistical literacy variable supports earlier models of the hierarchical nature of statistical thinking obtained from a larger pool of items.
Jane M. Watson, Ben A. Kelly and John F. Izard