Re: Comparing two variation coefficients


Michael.Lacy.junk@xxxxxxxxxxxxx wrote:
Reef Fish <Large_Nassau_Grouper@xxxxxxxxx> wrote:

Michael.Lacy.junk@xxxxxxxxxxxxx wrote:
Alfred <fredo@xxxxxxxx> wrote:
Dear All,

What test (possibly non -parametric) can be used to test
the differences between two variation coefficient (std dev / mean) ?

Thanks

Fredo

Here's a comprehensive source on your problem and related
ones.

Donnelly SM, Kramer A 1999. Testing for multiple species in fossil
samples: An evaluation and comparison of tests for equal relative
variation. American Journal of Physical Anthropology 108 (4): 507-529

I have to admit that this is the FIRST time that I have heard of
testing the equality two Coefficients of Variations, in parametric
or non-parametric statistics books.

Perhaps there are applications for which the test is appropriate for
a problem at hand, but I can't think of any at this moment.

Perhaps it would be more interesting and instructive for someone to
give a REASON for that specific test of difference of "coefficients of
variation" (a well-defined statistical term), without just citing some
article in a paper in a non statistical journal such as that of
Anthropology.

Actually, there is quite a bit of statistical material in the AJPA,
since physical (bio-) anthropologists are quite a different species
than cultural anthropologists. Of course, it is applied statistical
work.

You are far short of giving REASONS for that specific test of
difference of "coefficient of variation". To say that one kind of
anthropologists are quite different from another breed of
anthropologists in matters of applying statistics is about as
relevant as saying Brie is quite a different kind of cheese than
Camembert.

The article I mentioned is useful because it reviews
a lot of work from more "high-powered" statistical
journals, which I thought would present the material in a way that
could be useful to the OP. FWIW, the typical application of interest
among physical anthropologists is to compare the amount of variation
in some metric character, e.g. breadth of a particular molar,
between a sample from a known single-species reference population,
and an unknown fossil sample from a possibly multispecies collection.
(Excess variation in the unknown sample can suggest multispecies
composition of the unknown sample,

which could manifest itself in various known statistical tests of
MEANS,
VARIANCES, or various ANOVA methods, using both parametric
and nonparametric methods.

But why "coefficient of variation"? You hardly touched that
question!

a matter of some interest among
paleontologists and bio-anthropologists.)

-- Bob.

.

Relevant Pages

  • Re: Comparing two variation coefficients
    ... the differences between two variation coefficient? ... testing the equality two Coefficients of Variations, ... or non-parametric statistics books. ... since physical anthropologists are quite a different species ...
    (sci.stat.math)
  • Re: Varying coefficients in (non)linear regression
    ... > characterizes the variation. ... I know varying coefficient regression ... of modeling tools that you refer to, ... order models. ...
    (comp.soft-sys.matlab)
  • Re: Judaism, evolution, and science in general
    ... >>> of simple statistics. ... less likely to be *capable of* earning a living wage based upon their ... >>scores within these racial categories are much greater than the variation ... within each group renders that practically insignificant. ...
    (soc.culture.jewish.moderated)
  • Re: Judaism, evolution, and science in general
    ... >>any claim of statistical variation between racial ... > So it's like I said: there's good and bad racism. ... (which would be an abuse of statistics, ... >>Again, if you have a better label, then by all means use ...
    (soc.culture.jewish.moderated)
  • Re: Significance Test after Coefficient of Variation
    ... I am using coefficient of variation to describe the variability b/ ... takes CV into account instead of Standard Deviation, ... deviation of logis a measure of the CV, for small CVs. ...
    (sci.stat.math)