Talk:Parametric statistics

Latest comment: 4 years ago by Glenbarnett in topic History section somewhat misleading

Merge with Parametric model

edit

Please vote by putting your vote in bold.

  • No please don't merge it... Make a link from parametric statistics to it but parametric models don't necessarily have to be statistics and I wouldn't have found it otherwise... Thanks, M —Preceding unsigned comment added by 128.250.102.134 (talkcontribs)
  • dont merge - perhaps we could call it parametric 'model-ing' which is a very important term used to describe a type of object used in 3D modeling and rendering.
  • Merge. Yes, this should be merged with parametric statistics. However, I also would like you to define what "Parametric Study" too.
  • Merge. I'd also like to see the distinction between nonparametrics in the sense used in a phrase like "nonparametric regression" and what we might call "distributional nonparametrics" (the usual distribution-free stuff) disappear or be de-emphasized; they're not actually different things, but rather, different aspects of the same thing.

Move to category

edit

There is hardly enough here to warrant a page, thus I moved it to the category, much like I did with Category:Non-parametric statistics. I do not think this page is worth keeping separate. It provides scant information. This page will always be a stub here. — Chris53516 (Talk) 14:44, 13 February 2007 (UTC)Reply

Redirect from "Parametric analysis"

edit

I suggest a redirect from "Parametric analysis" to this article. There is already a corresponding redirect in place for "Nonparametric analysis".

Alopdahl (talk) 09:29, 15 March 2013 (UTC)Reply

Make it so. --Thosjleep (talk) 07:07, 16 March 2013 (UTC)Reply
It might be better to have disambiguation page titled something like "parametric analysis" pointing to here and other similarly titled articles ... parametric model, parametric equation, etc. some of which are not statistical. 81.98.35.149 (talk) 11:55, 16 March 2013 (UTC)Reply

Praise

edit

I find the article very well written and easy to understand. Thanks to the authors! Georg Stillfried (talk) 07:54, 13 August 2013 (UTC)Reply


History section somewhat misleading

edit

As far as I can see, Fisher (in the 1925 book referenced) uses the word "parametric" twice, both in the phrase "parametric function". He neither coined the term "parametric statistics" (that's Wolfowitz in the 40s in his PhD thesis, where nonparametric statistics is first defined), nor was he the first to consider parametric statistical models (not by a long way). He did some important work on both parametrics and nonparametrics but a bare reference to Fisher alone as if he somehow invented the concept is worse than not saying anything. At the very least offer an exact reference (preferably with a quote) so we understand exactly what is being claimed here. I don't think the text supports the current implication, and fixing it would make it not noteworthy enough to include. Glenbarnett (talk) 04:16, 15 January 2020 (UTC)Reply