Biology Reference
In-Depth Information
FIGURE 6.3
Difference in growth trajectories of the femur by year of age in Chilean and European children.
used elements to estimate stature. Ahandful of the studies in themeta-analysis look atmeasure-
ments of the cranium, face, or dentition, but with little success (r
5.0 cm) and so
should be avoided. Not surprisingly, the cranial sutures had the lowest success (r
<
0.06, SEE
>
0.09
e
0.363)
as there is no direct association between cranial sutures and living height! Multiple regression
yields higher results than simple linear regression. Stature formulae for males tend to have
higher correlations than those for females (Moore and Richter, 2012).
¼
What Method Should You Use?
Bioarchaeology
The problem with prehistoric populations is that no known contemporary reference
sample formulae are representative.
Sciulli (1990)
conducted a stature study for pre-Colum-
bian populations from the Ohio River valley. They used the original Fully method as their
actual stature on which they regressed the long bone lengths. They found that Trotter and
Gleser overestimated stature due to different body proportions of their population sample
(
Sciulli et al., 1990
).
Auerbach and Ruff (2010)
agreed with
Sciulli et al. (1990)
that the existing
regression equations were not appropriate for the diversity of Indigenous North American
populations. The designation of “Mongoloid” by Trotter and Gleser was not appropriate.
They therefore sampled 967 skeletons from 75 archaeological sites and created three
