3 ms·
http://pcl.missouri.edu/sites/default/files/freeLunch_0.pdf http://pcl.missouri.edu/sites/default/files/freeLunch_0.pdf has a neat example of why Bayes factors
by jamii 10y ago
http://pcl.missouri.edu/sites/default/files/freeLunch_0.pdf http://pcl.missouri.edu/sites/default/files/freeLunch_0.pdf has a neat example of why Bayes factors might be preferred over credible intervals.
> By using posterior credible intervals, we might reject the
null, but by using Bayes’ rule directly we see that this rejection is made prematurely as
there is no decrease in the plausibility of the zero point.
- stdbrouw 10y agoBut that's the thing, there is just never any real need to do hypothesis testing, the range of plausible effect sizes is all that matters.
- conjectures 10y agoI see what you're saying and mostly agree, but... This could totally come up if you have a non factorising multivariate posterior distribution, and you want some sound way to summarise it when you can't reason from the marginals alone.