Talk:Nuisance parameter

Frequentistic Version

This article still needs text relating to the frequentist point of view, and to link into ancillary statistic. May also need to rethink the redirect from "nuisance parameter". Melcombe (talk) 17:40, 5 March 2008 (UTC)

Merge Nuisance parameter and Nuisance variable

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.

I have removed the merge templates as there was no agreement after 1 year ...and there is a clear difference in meaning. Melcombe (talk) 14:45, 8 July 2010 (UTC)

---

I suggest (re-)merging Nuisance parameter and Nuisance variable

Both would sit well in an article beginning something like:

In statistics a nuisance parameter or nuisance variable is, respectively, a parameter or a random variable that is fundamental to a probabilistic model, but which is of no particular interest (or no current interest) in itself, yet must be taken into account in any analysis of the quantities which are of interest...

IMO it makes sense to consider the two notions together in one article, since they are so closely similiar; indeed, at least to a Bayesian they are entirely equivalent.

Also raised at WT:WPSTAT, to try to attract wider editor perspectives. Jheald (talk) 20:09, 30 June 2009 (UTC)

My response is that the articles should remain separate as the "nuisance parameter" interpretation of "nuisance variable" variable does not cover the other meanings of "nuisance variable" in the existing article nuisance variable. Both articles contain references to sources that use/define the two terms to have the distinct meanings stated in the articles. However there is as yet no source that uses "nuisance variable" to mean the same as "nuisance parameter" in any context and specifically not in a Bayesian context. All the sources I have just use "nuisance parameter" in Bayesian contexts. Melcombe (talk) 09:10, 1 July 2009 (UTC)

They are, nevertheless, exactly the same concept, and a Bayesian would deal with both in exactly the same way. We can note in passing that frequentists see a distinction between parameters and variables, but that is surely just a subsidiary note. When two notions are so closely related, it makes sense to treat them both together. (It also makes any presentation of claimed distinctions between them much easier and more transparent). You haven't talked to these points. For these reasons it makes more sense to treat the two ideas together, in the one article. Jheald (talk) 09:23, 1 July 2009 (UTC)

If the get merged, then nuisance (statistics) might be a suitable title. Michael Hardy (talk) 21:45, 1 July 2009 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

F-test variances are nuisance parameters?

Not sure whether this is a good example or valid. 68.134.243.51 (talk) 15:04, 5 August 2022 (UTC)

suggestion:Add a section on Likelihood-ratio test,score-tests and nuisance parameters

maybe link to the other pages also 68.134.243.51 (talk) 20:32, 5 August 2022 (UTC)