Stats/Methods theme news
Issued: Fri, 20 Jan 2017 11:49:00 GMT
We hope that the newsletters will provide an opportunity to hear from each of our themes throughout the year. Kicking us off are the Stats/Methods team with some details of their recent work:
Over the past year, members of the team have coordinated an effort to revamp undergraduate and postgraduate training in data analysis to emphasize reproducible research and open science.
One of the key changes in this effort was a step away from point-and-click, recipe-driven analysis and toward the generation of reproducible scripts using the programming language R. The team has also promoted active learning through the introduction of coursework across all levels, and filled a key gap in the curriculum by adding an advanced data analysis course for fourth-year undergraduates.
To bring new MSc students up to speed in R, the team conducted a three-day ‘R boot camp’ at the start of the academic year. In recognition of their efforts, the team was shortlisted in the 2016 Scotland Herald Higher Education Awards for the category of Enhancing Student Learning.
In addition to teaching, the stats/methods team has also organised activities aimed at improving scientific practices within the INP/School through the regular meetings of the ‘Statistics Tea Taste’ talk series. During the current academic year, we had a presentation from noted methodologist and open science advocate Dr Richard Morey from Cardiff University, who broke our intuitions about the meaning of confidence intervals. We also had vibrant discussions about treating stimuli as fixed versus random in neuroimaging and behavioural studies, and a workshop on specifying random effects in mixed-effects models.