Classification, Identification, and Individualization
133
of matching points or on analyst judgment (Ashbaugh, 1996; Cole, 1998;
1999). For genetic analysis, the population frequency of a profile, rather than
the
number of loci examined, is usually considered the more useful criterion.
Comparison of evidence and reference
Once the examiner has chosen potentially individualizing traits in one item,
she compares them to the other item. The examiner looks for patterns —
sizes, shapes, colors, or other measurements — that are indistinguishable
from each other. Traits are indistinguishable if they fall within the range of
variation that has been determined to exist for traits of that evidence type.
This is typically accomplished by repeated analyses of the same known trait.
For quantifiable data, the range of variation ideally reflects the results of
formal validation studies; for qualitative evidence, it
is often left to the expe-
rience and judgment of the individual analyst.
At the same time, the analyst must actively seek out differences between
the two items that might disqualify them as originating from a common
source (Smith et al., 1993). This is the heart of our attempt to rigorously
falsify the null hypothesis, that the items derive from a common source. The
analyst may decide that such occurrences are explainable or unexplainable,
and this decision will determine, in part, her conclusion. Even one
unex-
plainable
difference is sufficient to prove the null hypothesis and negate a
potential individualization.
Because of the
nature of forensic evidence, the task of deciding whether
traits are similar or dissimilar is less than straightforward. How hard should
we look? On what scale? At some level, every item will appear different from
its true source. These are some of the issues over which two qualified analysts
might reasonably disagree. Because any differences seen (significant or not)
may be inherent in the evidence (
intrinsic differences
) or may result from
external factors, such as the way the evidence is transferred or detected
(
extrinsic differences
), the analyst must not
only be cognizant of the
nature
of the evidence
, but the
nature of the test
. In fact, the factors most likely to
confound the analyst’s ability to determine the authenticity of similarities or
differences between evidence and reference traits are external to the evidence.
It is beyond the scope of this book to explore the specifics of these issues for
every type of evidence. We present a few representative examples here and
leave the reader to extend the underlying concepts to her own work.
Physical match evidence
. Consider two pieces of paper that share a border
of the same general shape (Figure 6.4). To the
naked eye, they appear to fit
together, and most would agree that they were once one. Upon closer exam-
ination with a hand magnifier, regions of apparent mismatch are detected at
various locations along the border. When the border is examined under a
high-power microscope, many discontinuities are apparent along the length
of the border. The examiner must first decide which scale of detection is
8127/frame/ch06 Page 133 Friday, July 21, 2000 11:47 AM
134
Principles and
Practice of Criminalistics
appropriate for the evidence. Note that the scale of detection is a factor
extrinsic to the actual evidence. For example, if experiments have previously
shown that two halves of the same piece of paper always appear noncomple-
mentary under high power, this scale of detection may be too high relative
to the scale of the evidence and the scale of the characteristics. Conversely,
if experiments have shown that two pieces of paper representing halves of
different sheets in the same stack torn simultaneously are indistinguishable
to the naked eye, the scale of detection may be too
low relative to the scale
of the evidence and the scale of the characteristics. Selection of the exami-
nation procedure depends both on previous experimentation and the expe-
rience of the field as a whole and the individual examiner. If the examiner
decides that discontinuities seen along the boundary of the paper fragments
are both real and significant, he must attempt to discern whether the expla-
nation is trivial or consequential. Two possibilities exist. Either the two pieces
were once one and conditions since their separation have conspired to
degrade or obliterate the detail along one or both edges. The
analyst might
adopt this explanation for any differences noted. Conversely, degradation
might fail to explain the differences (for example, if protrusions existed at
Do'stlaringiz bilan baham: 
