Statistical meaningfulness, teaching craft and writing about teaching statistics and data science

Abstract

In this issue we celebrate the awarding of three prizes for papers on good practice in teaching statistics and data science: the C Oswald George prize for best paper in Teaching Statistics issues 1 to 3 in 2021; the Peter Holmes prize for highlighting excellence in motivating practical classroom activity in these issues; and the Teaching Statistics Trust prize for best paper in the 2021 special issue on Teaching Data Science and Statistics. The announcements and citations for these three papers may be found in this issue. These papers significantly add to the demonstration of the work across the world in developing, implementing, sustaining and researching good practice in teaching statistics and data science, and this issue also includes editorial thanks and appreciation to all who contribute to the writing on such good practice authors, reviewers, the Teaching Statistics Trust and publisher Wiley. There is substantial need for, and considerable appreciation of, more writing of high standard on good practice in teaching statistics and data science. Full text downloads of papers in Teaching Statistics increased by approximately 30% from 2016 to 2020, but increased again by more than 33% in just the first 10 months of 2021 with the addition of the special issue. There is also need for a substantial cultural shift with greater acknowledgement and respect for the skills and expertise required for good teaching of statistics and data science in and across all disciplines, especially foundation and introductory, and for creditable and refereed writing on good practice in such teaching in its development, implementation, sustainment, evaluation and research. Hence there is need for understanding of both what constitutes good practice in teaching statistics and data science, and what constitutes good writing and researching such practice. Although there has been much discussion over the past three decades on the former, this discussion must be ongoing and constantly evolve to reflect the constantly evolving and dynamic nature of statistics and data science as they develop diverse capabilities (methodological, conceptual, and technological) to tackle increasingly complex and large problems in wideranging real contexts. Clearly the first requirement of the latter is that it must be about good practice in the teaching. However it should also satisfy criteria of scholarly writing but appropriate for the very large community of all those who teach statistics and data science. In the interests of reader convenience, I am now going to use the word “statistics” instead of “statistics and data science” to include everything to do with thought, endeavours and professional practice involving chance, variation and data, without attempting to describe any internal or external possible “boundaries”. There has been much emphasis over many decades that good statistics teaching must reflect the good practice of statistics, but the parallels between the two are far deeper and more complex than just a reflection. Recent and current data science fervour has led to an increase in speaking and writing by leading statisticians about the professional practice of statistics, often building on wise words from the past, such as Tukey [4], and also commenting on leadership in the profession and its practice. In researching leadership in statistics, in response to some recent requests, I was struck by the parallels between what statisticians wrote about leadership in the practice of statistics, and what I had observed over many years of what constitutes leadership in good practice in teaching statistics. Here, however, I want to focus on the challenges and critical importance of good writing about good practice in teaching statistics. To enable the ongoing development and, crucially, sustainable implementation of the latter we need much more of the former, and wider and higher acknowledgement of the former by higher I mean by higher authorities and by leaders. We instantly have the problem of this word “good”. The word “good” does not have to be used as much in speaking of research and research literature because standards and criteria relevant to each research field have grown and become established; if they are debated even if vehemently so the debate has a reference framework. This is similar to the principle that it is more productive to bring at least some framework of propositions or plans to a meeting. Most research journals “belong” to professional societies, and many in statistics have put considerable effort into work to develop and sustain applied sections of their journals, as well as high quality reviews and DOI: 10.1111/test.12300