Resumen de Improving learning in statistics – A pilot trial study to achieve the triple effects of depth, breadth and integration

Eric Shiu

• Statistics is an essential subject in management and science education. Yet the teaching and learning of this subject is often not smooth particularly for less numerical students. The conventional teaching approach, typified by students passively receiving statistical knowledge taught through the use of made-up data, has been discussed in previous statistics education research, and a number of revised approaches including problem-based learning, experiential learning, graphics and simulation, and understanding of students' previous statistical knowledge, have been put forward, all of which have the overriding aim at improving learning in statistics.

  All of these revised approaches have successfully shown the benefits of their focus on the depth of learning. However, these benefits do not come without costs, as most of them sacrifice the breadth of what students should effectively learn. What's more, it is apparent that none of these approaches nurture an environment encouraging students to integrate the statistical concepts they learn.

  To achieve improved learning in statistics, this study argues that deep learning constitutes only one facet of learning – breadth and integration being the others. Students should also have the right to learn the whole breadth of statistical knowledge as it was with the conventional teaching approach. They should also be given a learning setting within which they could engage in integrative learning.

  This paper concentrates on the pilot trial of a new teaching approach that aims at achieving the triple effects of depth, breadth, and integration in student learning of statistics. It will give an account of the theoretical underpinning, the procedures taken, and the evidence collected for the new approach with the potential to achieve these three effects. As the pilot trial has such an ambitious aim, it is likely there may be some further room to fine tune the new teaching approach. The author does not claim that the new approach introduced in this paper is an optimal solution for teaching statistics. Instead, the new teaching approach, after having achieved a successful pilot trial outcome, should be positioned as an approach showing signs of improvement compared to the existing ones