Data-based distributions of unknown parameters?
The very theoretical end of statistics deals with the assessment of uncertainty. How reliable are the conclusions from data showing appreciable haphazard variation? The issue has a long history essentially involving interplay between two notions of probability, one representing unexplained variability in empirical phenomena and the other representing uncertainty in knowledge. David Cox’s video, prepared for a recent Conference at Rutgers University, is partly about the general issues involved and partly about more specific and detailed issues. To watch the full video of the seminar follow the link: https://www.youtube.com/watch?v=0nXppvTPeo4&feature=youtu.be