Talk:Fiducial inference

I have written a short introduction to fiducial inference to start the ball rolling. It is not yet highly polished, but it's a start. I would sign my name, but after dozens of attempts to create a username that was accepted by the system, I gave up ("only takes a few seconds"--yeah right!)218.101.112.213 08:04, 3 November 2007 (UTC) On reading the entry, I find it needs a lot more work. I am now working on polishing it (and have now managed to find an acceptable user name).TerryM--re (talk) 21:39, 17 November 2007 (UTC)

Quotations should have quotation marks

The phrase "not a probability distribution in the usual sense" appears on page 156 of Stuart, Ord, and Arnold, which tries to update Kendall and Stuart. (I removed the "neither truthful nor 'good faith' in the usual sense" phrase.) Kiefer.Wolfowitz (talk) 20:59, 11 March 2010 (UTC)

"Obsolete" scientific theories

In the categories "obsolete statistical theories" or "obsolete statitistical procedures", the adjective "obsolete" was chosen for compatibility with the super-category "obsolete scientific theories". Would Melcombe and other editors suggest an appropriate term to replace "obsolete" --- e.g., would "widely deprecated" be acceptable? 20:24, 12 March 2010 (UTC)

Resurrection

I added the quote from Davison, in Cox's collection from Biometrika. Even the most renowned statisticians associated with the neo-Fisherian revival of the 1970s clasify efforts at resurrecting "fiducial inference" as a form of necromancy, if I may state my POV here. Kiefer.Wolfowitz (talk) 22:13, 12 March 2010 (UTC)

Original research tags

Melcombe's stating that a remake of Kendall's dictionary mumbles something about fiducial inference (under statistical inference) was original research. My stating that a better dictionary (also based on Kendall's) doesn't mention fiducial inference under statistical inference is also original research. Both statements should be deleted.

It is inappropriate that Kendall is being used as the main source for this article, since Kendall's reliability has waned, while the reliability of Cox and his students (Hinkley and Davison) have waxed, and they don't repeat Kendall's weird tics about the confusion between confidence intervals and "fiducial intervals". (Kendall's statements of the 1940s would have been appropriate sources for a 1940s Wikipedia.) Thanks, Kiefer.Wolfowitz (talk) 21:51, 23 March 2010 (UTC)

Obvious error

The formula

$P\left(X>{\frac {\omega }{a}}\right)=1-a^{n}=\alpha .$

is unattributed and clearly does not make sense. Based on the previous formula is should be

$P\left(X>\omega {a}\right)=1-a^{n}=\alpha .$

EcuPadic (talk) 14:57, 5 February 2016 (UTC)

totally incomprehensible to non expert

without a lot of fancy lingo, can someone write a paragraph taht explains the subject the article as writtenfails totally

sorry — Preceding unsigned comment added by 2601:192:4280:F0D0:419F:4CBE:7BA7:4986 (talk) 15:36, 22 May 2016 (UTC)