(Apologies for cross-posting)
The most recent issue of International Statistical Review [75(3)] is devoted to statistics education and begins with a themed suite of seven papers designed to have broad appeal to practitioners and technology buffs as well as statistics educators. Its focus is on how to give students experiences much closer to the practice of professional statisticians than has been possible in the past using technology to collapse the time scales needed for instruction via virtual environments. The main areas used as illustrations are the design of experiments and multivariate analysis. Although these are relatively advanced topics, many of the ideas discussed can be applied at any level of statistical education.
To introduce readers to the area, its enormous potential for the advancement of statistics education, and also to promote the other papers in the themed suite, the journal has made available free online:
"Virtual Environments and the Acceleration of Experiential Learning"by Chris Wild
The issue continues its educational theme with an update of Joan Garfield's classic 1995 paper "How students learn statistics", reviewing the subsequent fifteen years of research, and an article in which Hirokuni Tamura issues the following challenge. You have a brief opportunity at a fundraising dinner to convince an influential advisory board member and potential donor of your school of the importance of business statistics. How do you do it? And how do you ensure that the business statistics that is taught lives up to this billing?
DETAILS ABOUT ISR Vol. 75, Issue 3:
"Virtual Experiments and Their Use in Teaching Experimental Design" by Paul L. Darius, Kenneth M. Portier and Eddie Schrevens
Darius et al. lead off with a discussion of the issues involved in students learning to design experiments by actually designing and conducting a significant number of experiments themselves. This is in stark contrast to learning to analyse data from experiments designed and performed by others which is the reality of most experimental design courses. The technology that makes this achievable in a timeframe that is realistic within a single course is a series of virtual environments, such as a greenhouse, where students can formulate an experimental plan and lay out the experiment with a great deal of freedom of choice. The environment then generates the data.
"Dynamic, Interactive Documents for Teaching Statistical Practice"by Deborah Nolan and Duncan Temple Lang
Nolan & Temple Lang begin with a broad discussion of deficiencies in statistical education, and then offer a technology called "dynamic documents" that can help us overcome some of them. The setting in which their arguments are advanced is allowing students to experience of more of the complexity of the data analyses performed during a substantial investigation. Their illustrative document combines a rational linear presentation of the analysis of data that reveals some of "the answers" with an account of the journey, its turnings and wrong turnings, allows students to explore all of these pathways and even modify the analyses and head down new pathways. Dynamic documents include all the control features usual in dynamic content such as sliders, decision boxes and so on, but also live and modifiable computer code.The system that allows such documents to be produced efficiently can harness the full power of R for graphics and analysis. Although the paper has been written entirely in terms of statistics education, the system they are describing provides a potent infrastructure for communication in many other contexts as well. A likely area of application is empowering statistical consultants to produce sophisticated decision-support documents for their clients based upon the consultant's data analyses. Such documents would enable clients to explore, dynamically, very complex scenarios - varying environmental assumptions, projections, inputs and business strategies.
Other authors, from a wide variety of backgrounds but with expertise that touches aspects of Darius et al. and Nolan & Temple Lang, have been invited to use these papers as points of departure for papers drawing on their own experiences at the interfaces between statistics education, educational research and technology.
"Virtual Environments and the Acceleration of Experiential Learning" by Chris Wild
discusses the exciting new possibilities that are opened up by virtual environments together with cognitive and pedagogical imperatives to be addressed to ensure that environments actually do teach the lessons they were designed to teach.
"One Possible Frame for Thinking about Experiential Learning" by George W. Cobb
Cobb, a leading educator and author of an innovative book on experimental design, develops the themes of abstraction and transfer. This is done with particular reference to building the skills needed for the designing experiments in practice. However, the discussion has broad applicability to all of applied statistics.
"Computer-Aided Statistical Instruction-Multi-Mediocre Techno-Trash?"by Carl J. Schwarz
Drawing on his experience as a pioneer of virtual environments in statistics education, Schwarz discusses the design of statistical investigations and the utility of computer aided instruction broadly and asks, "What separates the gems from the trash?"
"On the Utility of E-Learning in Statistics"by Wolfgang Härdle, Sigbert Klinke and Uwe Ziegenhagen
Härdle, Klinke & Ziegenhagen have worked for well over ten years on a series of major initiatives that have linked electronic and dynamic content, statistical software and print publishing. They also have long experience with using the resulting products in teaching. Their paper describes these ventures and asks the hard question, "Has all the effort been worthwhile?" - reaching for the data available to them to try to answer it objectively. They also impart the lessons from their long experience of working in large-scale, collaborative software projects.
"Statistical Thinking in Computer-Based Learning Environments"by Gilberte Schuyten and Olivier Thas
In the final paper of themed series, Schuyten & Thas discuss the development of statistical thinking, lessons from a variety of research literatures on hypermedia learning environments, and their own experience of teaching using the environments of Darius et al., thus bringing the discussion full circle.
"How Students Learn Statistics Revisited: A Current Review of Research on Teaching and Learning Statistics" by Joan Garfield and Dani Ben-Zvi
Garfield & Ben-Zvi update one of the classic ISR statistics education papers, Joan Garfield's 1995 paper "How students learn statistics", reviewing the subsequent fifteen years of research.
"Foundational Value of Statistics Education for Management Curriculum" by Hirokuni Tamura
Tamura addresses the structural problems that business statistics curricula are facing and issues the following challenge. You have a brief opportunity at a fundraising dinner to convince an influential advisory board member and potential donor of your school of the importance of business statistics. How do you do it? And how do you ensure that the business statistics that is taught lives up to this billing?
The issue closes with reviews of 22 books
Adapted from the Preface,
International Statistical Review (2007), 75, 3, 279-280.