Should we redefine statistical significance?

Significance Tests: Rethinking the Controversy


Recently, a team of statisticians and scientists suggested redefining statistical significance to p<.005, primarily on Bayesian grounds (Benjamin et al, 2017). Their arguments are explored from both Frequentist and Bayesian perspectives; it will be suggested that they cannot be sustained from either.

Sep 5, 2018 11:20 AM
Cardiff, UK
Cathays Park, Cardiff, Wales, CF10 3ND