Talk:Statistical probability

From Wikipedia, the free encyclopedia

IMHO these points of view are not opposing each other ? Ericd

For the most part, this is a 'religous' issue (in the computing/usenet sense). The closer two groups, the more intolerable the differences between them appear. While the philosophical differences between Jaynes and de Finetti seem minor to the outsider, they differ markedly in technical detail and rationale, but probably less in ultimate utility, Jaynes also rejects much frequentist orthdoxy, but concedes that it is mostly the dark corners that give trouble - for the rest, it is the justification that is wrong. (FWIW i believe Jaynes to be the most right so far.) 193.116.20.220 12:38 20 May 2003 (UTC)

This article is problematic.

  1. It's inaccurate.
  2. It's mostly or entirely redundant with other articles about probability.

The name of this article, "Statistical probability", suggests that it will discuss a specific interpretation of probability (as with "epistemic", "aleatory", "personal", etc) but in fact this article is just a list of interpretations, and as such it's redundant with other articles that have the same list, such as Bayesian probability and Frequency probability.

To be specific about inaccuracies:

  1. "Due to the phenomenon known as Statistical Regularity" -- there is no such natural phenomenon; "regularity" is solely a mental construct.
  2. "relative frequency is universally accepted as a legitimate interpretation of probability" -- Bayesians don't, since the frequency definition involves infinite populations or infinite repetitions; this statement needs to be weakened or qualified -- Bayesians accept relative frequency as an approximation.
  3. "expected rational utility" -- Cox and Jaynes didn't bring utility into the picture; Jaynes explicitly separated the derivation of laws of probability from utility. In any event, "expected rational utility" is not a phrase I've ever seen before -- is it a conglomeration of "expected utility" and "rational expectation" or something?

Given the difficulties, my suggestion is to identify the content unique to this article and merge that into one of the other articles, perhaps probability or better still, probability interpretations. Comments? Wile E. Heresiarch 16:00, 27 Dec 2003 (UTC)

I think all that's fit to print should be moved to "probability interpretations" - that article makes a tripartite division which is unnecessary as far as I can tell. The difference between "statistical" and "philosophical" interpretations is hair-splittingly thin, so better to put the whole discussion in one place. Onebyone 01:58, 4 Jan 2004 (UTC)

From VfD

  • Statistical probability -- not a phrase used in the statistical or probabilistic literature. My basis for that assessment: having read some dozens of articles and books on probability and statistics topics; also, a Google search for "statistical probability" yields Wikipedia and Wikipedia copies among the top links, a number of accidental conjunctions of "statistical" and "probability", and nothing that could be called a basic source. The Wikipedia article itself doesn't bother to define "statistical probability" but rather only contains a collection of links to other articles. As such the article is unneeded, uninformative, and redundant. See also my comments on the talk page. Wile E. Heresiarch 16:00, 3 Jan 2004 (UTC)
    • Move content into probability interpretations and redirect there. Onebyone 02:01, 4 Jan 2004 (UTC)
      • I'd like to suggest that there be no redirect page: of the pages that link to Statistical probability, most are lists of the form "See also FOO, BAR, and Statistical Probability"; the only one that occurs in article text is a misdirected link -- specifically I mean the one in the final sentence of paragraph 5 of Multi-valued logic, which goes [[Statistical probability|probability logic]] -- that should just link to probability logic. So I think all links to Statistical probability can be erased; no need for redirection. Wile E. Heresiarch 04:35, 4 Jan 2004 (UTC)
    • Keep [n' copyedit; define "statistical probability"]. JDR
      • The point is that there is no definition to put there, because there is no such technical term "statistical probability". Wile E. Heresiarch 06:52, 9 Jan 2004 (UTC)
I find several mentions of it in Google. Such as "To study voids by means of statistical probabilities" from Annu. Rev. Astron. Astrophys. 1988. 26: 245-294 Rmhermen 22:18, Jan 9, 2004 (UTC)
        • OK, I withdraw the request for deletion. There appears to be enough nontechnical references to "statistical probability" to make the article a redirect. In looking at some documents indexed by Google, it seems the term is used informally to mean relative frequency or proportion. I will edit the page accordingly. Wile E. Heresiarch 22:53, 9 Jan 2004 (UTC)

[edit] Discussion