Teaching Applied Statistics in the Digital Wave

Applied statistics, also known as quantitative methods, often present themselves to non-specialists as a block, a fixed set of skills that would be consensual within the circle of specialists, and who would only evolve along a scale of increasing sophistication. On the same note, these skills would form a sequence that requires to be learnt in a unique order, from basic to advanced. However a review of textbooks, research publications and courses shows that there is more than one way to conceive of statistics and practice them, which are not necessarily contradictory with each other, but not plainly in agreement either. There is also more than one way to teach statistics, not only because teachers do not put the accent on the same concepts, methods or applications, but also because there may be several ways to teach the same method, all with their specific relevance, depending on the teacher’s teaching strategy and style, on the audience’s social and intellectual characteristics, and on the material and symbolic pedagogic context. This communication will start with a attempt to classify classical, 20th-century approaches to statistical teaching (and learning), using a few examples from simple and more complex topics. Second, it will analyse the impact of the digital wave onto higher education teaching at large, that is, among others: the tremendously increased availability and accessibility of data, methodological guidance and exercises; the dissemination and diversification of computing devices; the enhanced possibilities of horizontal and vertical interactions, including with distance in time and space; the proliferation of statistical displays and analyses in wide- and more specialised audience media; the revived debate about the relevance and transparency of public statistics. Third, it will evaluate the impact of these evolutions on the above-mentioned typology of statistical teaching approaches. Finally, it will propose a few ideas that could help make teaching applied statistics to social scientific audiences a less esoteric, more efficient and enjoyable experience than it sometimes appears to be (and is).