Biology Reference
In-Depth Information
part by these cases. One of the outcomes of the 2009 NAS report was a call for more outside
scientists to become involved in forensics, and for forensics workers to develop more con-
nections with the general scientific community. Given these current trends, and some of the
interesting questions posed in forensics, we chose to add a chapter on these topics.
In the forensic sciences, as in biology, anthropology and paleontology, there is much
interest in comparing the shapes of objects. There are several specific areas of forensics in
which morphometric methods, both traditional and geometric, have been applied. The first
is the biometry of human form, where morphometric methods have been used to identify
human remains and to estimate the age, sex and ethnic background of both pre- and ante-
mortem humans. They can also aid the diagnosis of developmental or cognitive disorders
which might be factors in sentencing. Non-invasive age estimation of living humans is cur-
rently of tremendous interest because immigrant and refugee policies in many countries,
differ according to the applicant's age. Juveniles are often given a different, typically more
advantageous status in consideration of refugee status or eligibility for immigration and
social services (
Solheim and Vonen, 2006; Crowley, 2007
). Age estimation may also be a
factor in criminal cases because sentencing guidelines are usually different for adults and
juveniles. The second application of morphometrics in forensics is in the area of impres-
sion and pattern evidence. Impression evidence is created when two objects come in con-
tact forcibly enough to create an image or disruption of the distribution of material on one
or both objects, recording their interaction. Examples include fingerprints, palm prints,
tool-marks and footprints.
In this chapter, we will first look at a morphometric approach that retains size informa-
tion in the analysis (Procrustes Size Preserving, or Procrustes-SP). Next we will discuss
what it means to say that shape data from different sources “match” one another; which
differs from the usual concerns of biological studies with differences or variation. We will
then examine three research programs that have used geometric morphometrics to address
forensic problems in different areas and which also introduce new conceptual material.
The first of these programs focuses on forensic bitemark analysis, primarily on the issue of
variability in the anterior dentition as it is recorded in a bitemark. The question is whether
this is a unique identifier of an individual. Studies directed at answering that question use
the Procrustes-SP and Procrustes-based matching procedures. The second research project
uses geometric morphometrics to estimate the sex of an individual post-mortem; we
compare a discriminant function analysis with the
k
-means method of identifying group
membership based on shape data. The third project applies three-dimensional geometric
morphometric analysis to examination of human brain scans for evidence of Fetal Alcohol
Spectrum Disorder (FASD) in a forensic context. The FASD studies illustrate the use of rel-
ative eigenanalysis to allow classification based on excess variation (hypervariability)
within the FASD population relative to the normal population, and the use of likelihood
ratios to report the strength of forensic evidence in the courtroom.
SIZE AND SHAPE
In most biological applications of morphometrics, there have been efforts to separate
size and shape, and often to analyze shape independently of size. However, within the
