OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

office-comment message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: Re: [office-comment] Re: Gaussian Distribution vs Normal Distribution


I strongly suggest considering some other statistical authorities as well.

I remember the late prof. Feinstein strongly advocating the use of *gaussian distribution* instead of the very misleading term normal distribution.

Persisting in erroneous namings is not a virtue, and IF many do it, it still is NO virtue, when better alternatives exist.

Also, as the newer statistical methods get into common use
(I refer here at bootstrapping procedures and other non-parametric
tests) the so called normal-distribution will loose most of its
appeal and usefulness. Actually, many professional statisticians
use less and less often tests based on the gaussian distribution.

I definitely hope that the t-test and the similar get replaced in the
next 5 to 10 years by the vastly more robust and more exact non-parametric
methods. Therefore, it is wise to specify exactly within a standard
the full name of the distribution.

I definitely recommend discussing these issues with statisticians on the
forefront of development. It will minimize a lot of pain in the future.


Leonard Mada

> (Regarding the OpenFormula specification):
> Leonard Mada:
> > I strongly suggest using within a standard document the more accurate
> and 
> > complete name:
> > *gaussian distribution* instead of the often misleading *normal 
> > distribution*
> Thanks for your comments.
> "Complete to who" is a problem, though.  A quick review of available
> documentation suggests to me that "Normal distribution" is the standard name,
> and that "Gaussian distribution" is the uncommon (nonstandard) term.
> Here's some pieces of evidence that convince me that "Normal distribution"
> is the more common term:
> * Wolfram's MathWorld uses as its primary term "Normal distribution", not
> "Gaussian distribution", and has this to say in
> http://mathworld.wolfram.com/NormalDistribution.html :
> 'While statisticians and mathematicians uniformly use the term "normal
> distribution" for this distribution, physicists sometimes call it a Gaussian
> distribution and, because of its curved flaring shape, social scientists
> refer to it as the "bell curve."'
> * It's worth noting that Wikipedia uses "Normal distribution" as its
> primary name; Wikipedia has a rule that articles should use the most common name
> for the article, giving more evidence that this is the right name.
> * A Googlefight (showing which term is more popular on the Internet) shows
> "Gaussian distribution" with 943,000 references while "normal
> distribution" gets 14,100,100 references:
> http://www.googlefight.com/index.php?lang=en_GB&word1=Gaussian+distribution&word2=Normal+distribution
> We should choose the terms that are more common, generally, so that we can
> communicate - and by that measure "Normal distribution" wins.
> In addition, since this is one of the _statistical_ functions in the
> formula spec, it seems appropriate to use the standard terminology used by
> statisticians. Wolfram's text in particular argues that the term should be
> "Normal distribution".
> While there are obviously other statistical distributions, I think the
> central limit theorem is a pretty good argument for NAMING this distribution
> the "normal" distribution.  This theorem states that "Under certain
> conditions (such as being independent and identically-distributed with finite
> variance), the sum of a large number of random variables is approximately
> normally distributed" [Wikipedia text, but this is well-known in
> mathematics/statistics].  Which means that when things get added up, even if they didn't
> start with a normal distribution, they converge towards it.
> The spec _should_ include the term "Gaussian distribution" when discussing
> this function - that's fair enough.  But it appears to me that the
> standard name for this is "Normal distribution" - the alternative terminology
> seems to be primarily in specialty areas (e.g., physics).  We should strive for
> the most common term, and if there isn't an obvious common term, use the
> term that the primary experts use (in this case statisticians).  Either way,
> I think "Normal distribution" wins.
> --- David A. Wheeler

Psssst! Schon vom neuen GMX MultiMessenger gehört?
Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]