A Longitudinal Study of Student Understanding of Chance and Data

This study uses Partial Credit Rasch analysis to study a complex data set of student responses to survey items relating to chance and data. The items were administered in the classroom and collected from 1993 to 2003 in the Australian state of Tasmania. Data were collected from a total of 5514 individual students across Grades 3 to 11 over the decade and of these students 896 provided at least one repeated measure. As students completed a core of common items, Rasch analysis could be performed and all students were subsequently placed on the same logit scale for comparison. The purpose of the analysis is to consider average cohort change over time and trends in performance during the first 10 years after the curriculum was introduced in Tasmania. Implications for the education system and curriculum implementation are considered.