fiducial on a string

A very short note in arXiv today by Gunnar Taraldsen and Bo Henry Lindqvist (NTU, Norway). With the above title. I find the note close to unreadable, I must say, as the notations are not all or well- defined. The problem starts from Teddy Seidenfeld [whom I met in Harvard around Dutch book arguments] arguing about the lack of unicity of fiducial distributions in a relatively simple setting. Actually the note is also inspired from Bayes, Fiducial and Frequentist, and comments from Teddy, a talk I apparently missed by taking a flight back home too early!

What I find surprising in this note is that the “fiducial on a string” is a conditional distribution on the parameter space restricted to a  curve, derived from the original fiducial distribution by a conditioning argument. Except that since the conditioning is on a set of measure zero, this conditional is not only not-unique, but it is completely undefined and arbitrary, since changing it does not modify the properties of the joint distribution.

Bayes is typically wrong…

In Harvard, this morning, Don Fraser gave a talk at the Bayesian, Fiducial, and Frequentist conference where he repeated [as shown by the above quote] the rather harsh criticisms on Bayesian inference he published last year in Statistical Science. And which I discussed a few days ago. The “wrongness” of Bayes starts with the completely arbitrary choice of the prior, which Don sees as unacceptable, and then increases because the credible regions are not confident regions, outside natural parameters from exponential families (Welch and Peers, 1963). And one-dimensional parameters using the profile likelihood (although I cannot find a proper definition of what the profile likelihood is in the paper, apparently a plug-in version that is not a genuine likelihood, hence somewhat falling under the same this-is-not-a-true-probability cleaver as the disputed Bayesian approach).

“I expect we’re all missing something, but I do not know what it is.” D.R. Cox, Statistical Science, 1994

And then Nancy Reid delivered a plenary lecture “Are we converging?” on the afternoon that compared most principles (including objective if not subjective Bayes) against different criteria, like consistency, nuisance elimination, calibration, meaning of probability, and so on.  In an highly analytic if pessimistic panorama. (The talk should be available on line at some point soon.)

principles or unprincipled?!

A lively and wide-ranging discussion during the Bayes, Fiducial, Frequentist conference was about whether or not we should look for principles. Someone mentioned Terry Speed (2016) claim that it does not help statistics in being principled. Against being efficient. Which gets quite close in my opinion to arguing in favour of a no-U-turn move to machine learning—which requires a significant amount of data to reach this efficiency, as Xiao-Li Meng mentioned—. The debate brought me back to my current running or droning argument on the need to accommodate [more] the difference between models and reality. Not throwing away statistics and models altogether, but developing assessments that are not fully chained to those models. While keeping probabilistic models to handle uncertainty. One pessimistic conclusion I drew from the discussion is that while we [as academic statisticians] may set principles and even teach our students how to run principled and ethical statistical analyses, there is not much we can do about the daily practice of users of statistics…

against the Dutch book argument

In continuation of the previous post, here are my slides for this afternoon talk at the 4th BFF conference (with nothing against Dutch people of course!, or anyone actually since this is a “best friends forever” conference):