Biology Reference
In-Depth Information
FIGURE 6.1
Statistical regression using the femur to predict stature.
where he measured 48 male cadavers, from which he developed regression equations. These
did not compare well with Pearson's European data and Pearson's equations did not work
successfully to estimate the stature of the Chinese individuals (
Stevenson, 1929
), indicating
that population-specific equations are needed.
Through the early twentieth century, the mathematical methods of the Manouvrier tables
and Pearson's equations became the standards for stature estimation. In 1948, Stewart recom-
mended better equations and testing using the large Hamann
e
Todd cadaver sample from
Case Western Reserve University in Cleveland, Ohio (
Stewart, 1979
). Two researchers inde-
pendently embarked upon the challenge: Wesley Dupertuis and Mildred Trotter. In tests of
the widespread applicability of Pearson's equations,
Dupertuis and Hadden (1951)
claimed
that the population Pearson used was short and should not be used on tall populations. Their
equations yielded stature estimates that were 8.5 cm taller than from Pearson's equations.
They also did not think there was a difference between living and cadaver stature, but Trotter
and Gleser disagreed (1952).
Dupertuis and Hadden (1951)
had produced “general equations” for use on any popula-
tion, despite the lessons learned from Stevenson that population-specific equations are
