Biology Reference
In-Depth Information
necessary (
Stevenson, 1929
). They recommended using more than two bones to estimate
stature (i.e., multiple regression) and that the lower limbs were better than the upper.
Trotter, on the other hand, echoed Pearson's recommendation for a representative sample
from the same sex, age, ancestry, geographic area, and time period. Trotter and Gleser's
formulae (1952, 1958) became more widely accepted than those of
Dupertuis and Hadden
(1951)
because they used a larger sample of American World War II soldiers (including Asian
data). They also provided a correction for age (
Trotter and Gleser, 1951a
). Trotter and Gleser
altered Pearson's method in three ways. First, they used a combination of living stature and
dry bones. Second, they introduced the first correction for age. Third, they
tested
the validity
on a second sample of individuals who died during the KoreanWar (
Trotter and Gleser, 1952,
1958
).
Population-Specific Regression Formulae
From the mathematical model for the estimation of stature that was gaining acceptance in
the field, it was becoming clear that population-specific data were needed. The study by
Trotter and Gleser (1952)
recognized some distinctions in limb proportion between African
Americans and European Americans.
Trotter and Gleser (1958)
tested their formulae against
data from the Korean War dead, and included new equations for Asian Americans and
Mexican males. They concluded that, due to the different limb proportions, different equa-
tions are necessary for each population. Their results indicated that stature estimation for
Puerto Ricans can be successful using equations for African Americans, but Mexicans
differed from all groups, so separate equations were needed. The standard error of estimate
(SEE)
3
was different for Korean War Europeans than for the WWII and Terry collections.
Further, Trotter andGleser analyzed the longitudinal data compiled byDr. Russell Newman
in 1950 with the U.S. Army Quartermaster Research and Development Command of 48,685
records (
Trotter and Gleser, 1958
). From this, they concluded that the age for achieving
maximum height had increased from the WWII dead (18 years) to the Korean War dead (full
stature at 21
e
23 years)
4
(see discussion later in this chapter). They measured all of the long
bones: humerus, radius, ulna, femur, tibia, and fibula. Trotter and Gleser determined that
the lower limb is more reliable than the upper, so there is no reason to use the upper limb if
the lower is present. They also recommended using the average of bone length of right and
left sides if both sides are completely present (
Trotter and Gleser, 1958
). See
Box 6.2
for lessons
learned from Trotter and Gleser's research about what not to do in stature estimation.
Although the work of Trotter and Gleser has been used and respected by researchers
around the world, one very serious error in their methodology has come to light.
Jantz
(1992)
tested Trotter's and Gleser's female formulae using data from the Forensic Databank.
5
They found that for European Americans, the femur and tibia yielded stature estimates that
differed by up to 3 cm between the bones, which led to further investigation of the Trotter and
Gleser methods.
Jantz and colleagues (1995)
made an important discovery that Trotter had
3
The standard error of estimate is used in regression analysis to examine how well a least squares line
equation fits a data set.
4
This secular change in the timing of growth is yet another reason why it is necessary to continually update
stature formulae.
5
This contains data from thousands of forensic cases curated by the University of Tennessee.
