Talk:P–P plot

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Statistics (Rated Stub-class, Low-importance)
WikiProject icon

This article is within the scope of the WikiProject Statistics, a collaborative effort to improve the coverage of statistics on Wikipedia. If you would like to participate, please visit the project page or join the discussion.

Stub-Class article Stub  This article has been rated as Stub-Class on the quality scale.
 Low  This article has been rated as Low-importance on the importance scale.
 

Removal of prod[edit]

A proposal to delete was put in, which I am removing. For info, the proposal said:

"concern = It has been shown in Makkonen L. (2008), Bringing closure to the plotting position controversy. Communications in Statistics - Theory and Methods, 37: 460-467 that the true rank probability is the order statistic mean P(i) = i/(n+1). Therefore, in a "probability plot", used in the way explained on the page, order statistics medians cannot be used. They will plot data at incorrect probability, and such a plot is useless. The error is particularly significant at the tails of a distribution.

The method explained on this page is, therefore, incorrect. In principle, the page could be corrected. However, since the same incorrect method appears in many other sources on this same title "probability plot" this might be misleading. For the time being, it would be better to delete this page and combine it with the page "Q-Q plot" (after correcting that page which is presently "disputed").|month = January|day = 8|year = 2009|time = 12:37|timestamp = 20090108123728 "

The article details ways in which probability plots have and will continue to be constructed and the potential fact that someone has published an article that may or may not purport to show that there is some particulr method that is better than others according to some possibily irrelevant criterion in no way affects this. This is not a text book nor something that only give details of someone's oppinion on what is best. Melcombe (talk) 13:14, 8 January 2009 (UTC)

REBUTTAL to Melcombe: In mathematics one should not consider the difference between correct and incorrect as "some possibly irrelevant criterion". The method presented on this page is simply wrong. It has unfortunately been used before, but history is not a good reason for using a wrong method. It makes no sense to assess whether or not a data set follows a given distribution if the data are plotted at wrong probabilities. The true probability is i/(n+1), as shown in the reference.

Recall:
  • Wikipedia is not a text book
  • Wikipedia is not a statement of recommended practice.
The article lists some formulae that people have used and continue to use. The reference you quote is somebody's conclusion in one particular context which certainly doen't cover all possible uses of plotting positions. The present article and the related others on Wikipedia may well need some more references and it might be appropriate to reference the one you quote also, but don't attach too much importance to it ... after all there have been studies of plotting positions for many years, none of which provide evidence of a single formula appropriate to all possible contexts. Melcombe (talk) 12:03, 15 January 2009 (UTC)