Biology Reference
In-Depth Information
landmark to the next was measured to determine the segment length and the percentage of
the total bone length. This research was carried out on a sample of Native American remains
from Arkansas using ordinary least-squares regression to get the best possible line to esti-
mate the total length of each bone. This could then be used to estimate stature by applying
Trotter and Gleser (1958)
and/or Genov
´
s (1967) stature formulae.
Steele (1970)
later presented formulae to estimate stature directly from each of the five to
six segments, skipping over the step of estimating the length of each individual bone. Each of
the bone segments is then regressed on stature using least-squares regression. These equa-
tions have a higher standard of error than when using a complete bone, but can be applied
if whole bones are unavailable.
Steele (1970)
also provided age correction formulae.
Bass (1979)
complained about the difficulty and inconsistency in finding the landmarks as
described by
Steele and McKern (1969)
. To address this concern and improve interobserver
error,
Simmons et al. (1990)
developed a newmethod for stature estimation from fragmentary
femora. They used Martin's measurements from 1957 for the femur (these are the same
measurements listed in Buikstra and Ubelaker, 1994), but reduced the number of segments
from 16 down to 9. The goal was to estimate the maximum length from each segment and
to determine stature directly from the segment. It is more accurate to eliminate the middle
step to estimate stature.
Simmons et al. (1990)
achieved means and standard deviations close
to those of
Trotter and Gleser (1952)
and better than those of
Steele and McKern (1969)
due to
a larger sample size. These revised measures by
Simmons et al. (1990)
are listed in
Box 6.7
; see
also
Figure 6.2
.
Bidmos (2008)
estimated stature using fragmentary femora of indigenous South Africans.
He used 100 skeletons (50 male, 50 female) from the Raymond Dart Collection to create a cali-
bration sample and then another 20 individuals as a test sample. Actual stature was calcu-
lated using the Fully method (
Fully, 1956
). He used only six measurements of the femur,
along with maximum length, four of which were the same as those from
Simmons et al.
(1990)
, namely: vertical neck diameter (VND), upper breadth of femur (VHA), bicondylar
breadth (BCB), and epicondylar breadth (EpB). He added medial condylar length (MCL)
and lateral condylar length (LCL) because these two measures are highly reproducible and
highly dimorphic. Ten centimeters were added to the stature calculations for “total skeletal
height” to account for soft tissues. The resulting standard errors of estimate were only
BOX 6.7
BOX 6.7
SIMMONS ET AL. (1990)
REVISED MEASUREMENTS
SIMMONS ET AL. (1990)
REVISED MEASUREMENTS
FOR FRAGMENTARY FEMORA
Maximum femoral length (FML)
Bicondylar breadth (BCB)
Vertical femoral head diameter (VHD)
Epicondylar breadth (EpB)
Vertical femoral neck diameter (VND)
Lateral condyle height (LCH)
Upper breadth of the femur (VHA)
Medial condyle height (MCH)
Transverse diameter of the midshaft
(WSD)
