Biology Reference
In-Depth Information
system. This method was first adopted for males (
Katz and Suchey, 1986
) and later for
females (
Brooks and Suchey, 1990
).
Brooks and Suchey (1990)
reported that previous research
(
Todd, 1921; Gilbert and McKern, 1973
) justified the need for a separate set of standards for
females due to shape differences and pregnancy-related changes in the pelvis.
The mean, standard deviation, and a 95% confidence interval are reported for each phase
for males and females. The 95% confidence interval is used as the predicted age range for
a given unknown individual and is therefore more appropriately called a “prediction”
interval (the term “confidence interval” has a different statistical meaning that does not apply
here, strictly speaking).
6
Because of their large sample, detailed phase descriptions, and
availability of corresponding casts,
7
the Suchey
e
Brooks method is the most widely used
method today for aging the pubic symphysis (
Figure 3.2
).
Most forensic anthropologists, after scoring the pubic symphysis phase (I
e
VI), would
report the mean of that phase as the “point estimate” for age-at-death for that individual,
along with the provided ranges on either side of that point estimate (
Garvin and Passalacqua,
2012
). A few forensic anthropologists have pointed out statistical problems with this kind of
reporting and instead advocate a more complex procedure called transition analysis (dis-
cussed earlier) (Boldsen at al., 2002) that can be used for any age-at-death indicator scored
in ordinal phases (see DiGangi and Moore [Chapter 1], this volume) (e.g., pubic symphysis,
auricular surface, sternal rib ends).
Studies continue on the pubic symphysis and the Suchey
e
Brooks method. For example,
Berg (2008)
reassessed this method for a modern American and modern Balkan sample and
using transition analysis, added a seventh stage as well as a redefined stage V and VI. More
recently,
Hartnett (2010)
found significant interobserver error with scoring (indicating
improvements in training or phase descriptions are necessary) on a modern sample of autop-
sied individuals fromArizona and also described a phase VII for individuals in their seventies.
It is clear that more validations should be done on this method for different populations.
The Auricular Surface
Lovejoy et al. (1985a)
were the first to propose using the auricular surface of the ilium as an
indicator of age-at-death. They had noted a high correlation between other skeletal age indi-
cators and morphological changes of the auricular surface. The auricular surface is of great
importance because it is more durable than other skeletal elements used for aging, and its
morphology does not appear to be affected by sex or ancestry (
Osborne et al., 2004
). As
with pubic symphyseal studies,
Lovejoy et al. (1985a)
stress the importance of understanding
human aging as a process. They recommend eight phases with corresponding age ranges
based on observations of morphological changes of the auricular surface. The first seven
ranges are narrow, each only encompassing 5 years, while the final phase encompasses all
individuals 60 years and older (see
Figure 3.3)
. This method can be applied to either males
6
Confidence intervals are 95% intervals of the distribution around a mean. An age-at-death estimate for an
individual is not an estimate of a mean, it is the estimate of a single observation from a distribution and so
involves much less confidence. The 95% intervals around a single observation are therefore referred to as
“prediction intervals” and are much wider than confidence intervals.
7
Casts are available from France Casting,
www.francecasts.com
.
