This paper is from Session 3C: Statistical instructors’ knowledge of assessing students’ learning of statistics

which comes under Topic 3: Education and development of staff who teach statistics

Discerning students’ statistical thinking: a researcher’s perspective

Presenter

• James Baglin (RMIT University, Australia)

Abstract

In the most comprehensive treatment of the topic to date, Wild and Pfannkuch (1999) propose a structural overview of the various domains of what constitutes statistical thinking. This paradigm provides a useful framework for approaching statistical thinking assessment. However, translating this model into assessments that are practical, reliable and valid for the purpose of statistics education research remains a challenge. This paper discusses the challenge of assessing statistical thinking and the development of a preliminary research-based assessment task based on Wild and Pfannkuch’s paradigm. Data collected from a large introductory statistics course where students completed the preliminary task are critically evaluated. Suggestions for future improvements to the task and ideas for alternative methods are raised. Statistical thinking may very well prove to be as difficult to assess as it is to define, but without further research, our understanding of how people learn to think statistically will be limited.