Path: utzoo!attcan!utgpu!watserv1!watmath!uunet!bionet!ARSUN.ANTHRO.UTAH.EDU!rogers
From: rogers@ARSUN.ANTHRO.UTAH.EDU (Alan R. Rogers)
Newsgroups: bionet.population-bio
Subject: Re:  help needed for probability scaling
Message-ID: <9009091524.AA21748@arsun.anthro.utah.edu>
Date: 9 Sep 90 15:24:56 GMT
Sender: daemon@genbank.bio.net
Lines: 28

From rogers Fri Sep  7 10:33:49 1990
To: SM01@liverpool.ac.uk
Subject: Re:  help needed for probability scaling
Status: RO

>Has anybody information about the probability scale (which was used in
>biology by Harding 1946 to dissect normal populations). I can't use
>probability paper b ecause of a huge data. What I need is a computer program
>which enables me to us e as probability paper on the computer or if I could
>know the formula which is used to produce probability scale. Thanks in
>advance. Ahmet.<sm01@liverpool.ac.  u.k>.

You are looking for what is called a "Quantile-quantile plot" in modern texts
on data analysis.  I suspect that it is now available in most statistical
packages.  I've seen it in Minitab, and use it all the time in the S
statistical package.  In S, the command you want is "qqnorm".  For example,

S> random.data <- rnorm(30)  # draw 30 standard normal variates
S> plot(qqnorm(random.data)) # plot quantiles of data against quantiles of
                             # Normal distribution

plots a nearly linear scatter of dots.  Consult pages 70-74 of "The New S
Language", by Richard A. Becker, John M. Chambers, and Allan R. Wilks.

Alan Rogers
 INTERNET: rogers@anthro.utah.edu
 USMAIL  : Dept. of Anthropology, Univ. of Utah, S.L.C., UT 84112
 PHONE   : (801) 581-5529