Teaching and Communication Tools for Enabling Statistical Thinking
5:00–9:00 PM CET
Many failings in the application of statistics do not result from errors in calculations but from logical and conceptual misunderstandings. Selection of methods for data analysis, interpretation of data and testing results, and application of statistical information in decision-making, for example, all demand strong foundations in probabilistic thinking. This course
supplies tools and best practices for teaching and communicating statistical thinking as a foundation for conducting, understanding, and applying statistics. While the statistical community has a wide variety of successful tools for teaching statistical calculations and tests, examples and resources for teaching statistical thinking are less common. We include examples from scientific studies, official statistics, working with decision-makers, and communicating to the general public.
The first half of the course will focus on common errors in statistical thinking and how to avoid them. It includes methods for thinking in simulations; how a single statistical analysis fits into the larger supply-chain of knowledge; and what p-values are and how they are often misused. We will conduct hands-on activities that can be used in a wide variety of settings, including on-line workshops. A set of these activities have paired R-labs for those teaching statistics courses.
The second half of the course focuses on statistical communication and the responsibilities of statisticians in communication of data and uncertainty. We cover some classic and fascinating examples of statistical miscommunication and work in groups on statistical communication challenges. Activities focus on communication of scientific results as well as official statistics and, in particular, on best practices for communicating uncertainty so that the uncertainty can
be leveraged in decision-making. These activities will be enable anyone teaching statistics as well as anyone consulting or communicating in the field of statistics to more clearly articulate the interpretation of scientific studies as well as why there is uncertainty, what it means, and how that information can be applied in a decision-making context.
Course materials that will be provided to participants include the course slides and lesson plans for both hands-on activities and R-labs as well as a reading list.
The target audience focuses on statisticians teaching statistics courses at the undergraduate and graduate level and will also be relevant to anyone conducting statistical consulting or communicating statistical work to those without a strong statistical background. There are prerequisites.