Dear
all
First off I
wish to thank those who replied to my plea. Have found the references
interesting.
Secondly,
interestingly enough even though this topic of measurement error (precision
and accuracy of measurements) is very topical in the world of anthropology, it
largely is either ignored or presumed to be low in the area of systematic
zoology and morphometrics in general. This I found most suprising as the
degree of measurement error has such influence on subsequent univariate and
multivariate analyses. Such that the resulting intra or inter- variation
obtained in morphometrics may in fact be due to the spurious effect of
measurement error rather than the actual difference or natural
variation.
In most cases,
the error is persumed to be low, but nevertheless it should be checked, even
if the measurer is experienced in taking the measurements. In my own case, I
decided on the following method after much trawling through the literature,
perhaps it will be of use for others.....? Obviously it is not the only way do
calculate the measurement error, rather just one thrown into the soup of
morphometrics for us to digest!
I have been
measuring various postcranial cervid bones for the last 5 years, and had to
check if my measurement values were accurate and precise on repeated measures.
Following
Lynch, J. M. and O'Sullivan, W. M. (1993), method of checking for measurement
error - they used the following equation (it uses the within-specimen
cofficients of variation calculated for each measurement, afterwhich these
coefficients are corrected for small sample size (Haldane, J.B.S. 1955 cited
in Lynch & O'Sullivan 1993):
ME% = 100
[(1+1/4n)SD/mean)]
where n
is the sample size (of replicates)
SD is the
standard deviation
mean is the
mean of the replicates per individual.
Lynch, J. M.
and O'Sullivan, W. M. (1993). Cranial form and sexual dimorphism in the Irish
otter Lutra lutra L. Biology and Environment: Proceedings of the
Royal Irish Academy 93B: 97-105.
Haldane, J.B.S.
(1955). The measurement of variation. Evolution 9:484.
Thanks again to
those who responded to my earlier plea, much appreciated.
Hope this email
finds ye all well
Kind regards
& Slainte
Ruth
--------------------------------------
Ruth F. Carden
B.Sc.
Mammal Research
Group
Zoology
Department
National
University of Ireland, Dublin
Belfield,
Dublin 4
Ireland
Tel:
(+)353-1-7162261 (note new number)
Fax:
(+)353-1-7061152
