Associate Professor in Statistics, Plymouth University
Paul Hewson started his career as a cell biologist in Scotland. He moved back to Devon in 1996, working in local government while retraining part-time as a statistician, and spending four of those years as a Road Safety Data Analyst and Research Officer at Devon County Council. Paul was awarded a PhD in 2005 and retains a particular interest in the role of Bayesian reasoning in understanding the world.
He was awarded a prestigious National Teaching Fellowship by the Higher Education Academy reflecting success in communicating with non-specialist statisticians; he regularly receives invitations to teach or examine around the UK and further afield.
He has continued to research topics related to road injury with a wide range of partners, and in 2016 was appointed an Associate Editor of the journal Accident Analysis and Prevention. In 2005, in partnership with Devon County Council and supported by a grant from the Rees Jeffreys Road Fund, he started work promoting evidence based practice training among road safety practitioners.
Presentation: Telling stories without spinning yarns: how to do data analysis in the 21st Century
With an over-abundance of data, it is now easy to find data to fit a story. Tyler Viglen has even set up a website to parody how easily we can find correlation in obviously unrelated variables. The thought processes that lead us into erroneous conclusions are as old as human thinking (see for example the poster here) and are not magically fixed by advanced mathematical statistical or database training.
This session will outline a process for interrogating data using relevant examples from social survey and social media. It will highlight those aspects of the story telling process that are core in avoiding gross interpretative errors. All users of data analysis (whether an analyst working with raw data, a practitioner reading a technical report or a citizen reading a newspaper newspaper article) should understand why a valid story telling process is core to obtaining credible insights from data.
This session will briefly highlight the potential to embed some road casualty reduction stories within the data investigation elements of the new Maths A level curriculum (for at least two of the exam boards).