[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Respiratory turbinates - again

From: endocrin@media.net.au (Graeme Worth)
 > ... What we then are left with is
 > the 2 lines for birds/mammals and reptiles. These appear to the eye to be
 > separate lines and may well be so, but the statistical assurance is missing.
 > I am not a statistician, but I'll bet there is a test that will demonstrate
 > the difference between 2 such lines at a particular degree of confidence.

There are several such tests (depending on what model one uses).

But all of them are based on the extent and type of overlap in values,
and thus require a good, robust estimate of *variation* within each
tentative group.  Hence my continual harping about sample size.

Try computing the standard error of the estimate of the variance
for a sample size of 12!!  Now compare that to the observed variation.
My educated guesstemate is that the standard error will be about the
same order of magnitude as the observed variance, or maybe even larger
than that.

 > I don't think that this translates as "dinosaurs were not endotherms".

True, but it reads to me as what I call "scientific caution".  That
is, in an intorductory paper on a new idea, the conclusion is
deliberately understated.  Then, if the idea doesn't pan out the
paper is still valid.

swf@elsegundoca.ncr.com         sarima@ix.netcom.com

The peace of God be with you.