Biology Reference
In-Depth Information
returning the orientation of the knee along the sagittal plane (
Mahfouz et al., 2007b
). This
latter three-dimensional study slightly contradicts the previous study (
Asala et al., 2004
)in
terms of distal changes of the femur being more significant than proximal end changes.
One reason may be that the three-dimensional analysis is likely more sensitive to the subtle
shape changes, because it compares thousands of homologous points as opposed to just a
few measurements with calipers. The results could be reflecting population variation in
the shape of the femur.
PATELLA
The patella has had some success in terms of sex estimation.
Kemkes-Grottenthaler (2005)
and
Mahfouz et al. (2007a)
looked at sex differences in the patella and found significant sex
differences in the maximum height and width of this bone. Introna and colleagues (1998)
yielded 84% accuracy using multivariate discriminant function analysis of both maximum
width and maximum height. Bidmos and colleagues (2005) had similar success in a South
African sample, with better success for height (85%) compared to breadth (79%). A more
complex method was used by Mahfouz and colleagues (2007a) that compared the three-
dimensional shape from computer models developed using CT scans of modern individuals.
This method achieved high success rates from a multivariate statistical analysis called fuzzy
clustering, with testing accuracy rates as high as 93%.
TALUS AND CALCANEUS
Gualdi-Russo (2007)
used discriminant function analysis to estimate sex from the talus
and calcaneus in a northern Italian population. The author tested the method on an indepen-
dent sample of 16 individuals that, despite a small sample size, had accuracy ranging from
87.9% to 95.7%, with the best accuracy achieved through a multivariate analysis of the talus.
The accuracy was only 56.3%when tested on a sample of Italians from southern Italy, empha-
sizing the need for population-specific data (
Gualdi-Russo, 2007
).
Bidmos and colleagues measured the sexual dimorphism of the calcaneus and talus of
South African populations and yielded similar results (
Bidmos and Asala, 2004; Bidmos
and Dayal, 2004
). The calcaneus alone achieved accuracy of 85% when using the three most
discriminating variables combined (
Bidmos and Asala, 2004
). Bidmos and Dayal also
measured the talus of South African Blacks and achieved accuracy rates of 89.2% when three
traits were combined (
Bidmos andDayal, 2004
). The best single trait was the height of the talus
head, which had an accuracy rate of 86.7%. When they tested this method using independent
samples of South African Whites and Blacks, the accuracy for the more comparable popula-
tion of other South African Blacks was high (70
e
90%), but the accuracy for the South African
Whites was much lower (40
80%). The authors argued that this is further evidence of popu-
lation specificity in sexually dimorphic skeletal dimensions (
Bidmos and Dayal, 2004
).
e
PROBLEMATIC AREAS OF SEX ESTIMATION
Subadult Sex Estimation
Subadult sex estimation has been perhaps the most problematic area of sex estimation for
both forensic anthropologists and bioarchaeologists. According to
Scheuer and Black (2004)
,
