266
Principles and Practice of Criminalistics
The Simpson Saga — The Blood
in the Bronco
On June 12, 1994 Nicole Brown (NB) and Ronald Goldman (RG) were brutally murdered. O. J. Simpson
(OS) was charged with
the commission of the crime, and among the evidentiary items confiscated was
his white Ford Bronco. A bloody smear on the passenger side of the center console was initially noted
and sampled on June 14, 1994. At least three relevant forensic questions can be asked about this evidence.
1.
How was the stain deposited?
2.
When was the stain deposited?
3.
Who may have contributed to
this stain?
For the purposes of this exercise we will
not attempt to address how and when the
stain was deposited. We will concentrate
on what the evidence can tell us about
who contributed to the stain.
DNA testing was used to investigate
the source(s) of the stain. Both DQ
α
and
D1S80 tests were performed on the two
evidence samples collected in June 1994.
From
the results, it was concluded that
one stain included Simpson as a donor, while excluding NB and RG. The other showed a mixture
consistent with OS and RG. A larger sample was collected on September 1, 1994. It should be noted that
in the intervening time period, the Bronco was burglarized. Although the items taken were themselves
of no evidentiary value, the integrity of the bloodstain evidence could no longer be guaranteed. Three
separate swatches were collected that more nearly covered the large area of the smear. These samples
were also analyzed for DQ
α
and D1S80. Let’s examine the data from two of these swatches in detail.
8127/frame/ch10 Page 266 Friday, July 21, 2000 11:38 AM
Good Forensic Practice — Obligations
of the Analyst
267
Above is a set of DQ
α
strips showing the results from two of the later samplings, along with their
substrate samples. (A substrate sample is collected from an apparently clean area near an evidentiary
stain. It is a way of assessing which, if any, genetic types might be present in the background.) One of
the ways an analyst avoids even an unconscious bias is to analyze the evidence before comparing it with
the reference samples. In this case, both evidence strips show the same pattern of dots, and even the
intensities are similar. In both samples, all
the dots, with the exception of 2 and 3, are positive, so we
can safely eliminate anyone with a 2 or 3 allele from having contributed to this sample. The next step is
to note that more than two alleles are manifest. Because any normal individual has, at most, two alleles
at any one genetic locus, this is a clear indication of multiple contributors.
From an examination of the nominal dots to the left of “C,” we note the presence a 4 allele and a
1 allele. The 1.1 and 1.3 dots are both positive and, although the 1.3 dot is substantially lighter, both are
stronger than the “C” dot. The 1.1 and 1.3 subtypes, at least, are represented. Is the 1.2 allele also present?
Consider the more difficult “trio” and “all but 1.3” dots, both of which are positive and greater than “C.”
The “trio” dot may be positive due to the presence of the 1.2 allele, the 1.3 allele, or the 4 allele
individually,
or any combination of them. Because we can confirm the presence of both the 4 allele and the 1.3 allele,
the “trio” would be positive regardless of the presence of a 1.2 allele, and so cannot be used to determine
its presence. This consequently makes the “all but 1.3” dot useless in determining, along with the “trio,”
the presence of a 1.2 allele. In short, from the pattern of dots on these strips, it is impossible to tell if a
1.2 allele is present in the sample or not.
At this point in the interpretation, the analyst would normally draw up a chart of alleles excluded
(2, 3), those positively present (1.1, 1.3, 4), and those about which we have insufficient information to
determine presence or absence (1.2). The analyst would also list possible pairwise associations of the
alleles into genotypes of the possible contributors (we will spare you this exercise). Enumeration of the
types in the DQ
α
system depends in part on compound dots which together determine a type. Because
of this, the analysis of mixtures becomes a bit complex. In addition, a mixture of bloodstains is often
more difficult to interpret
than a sexual assault mixture, in which at least one of the contributing types
(the victim’s) is often known.
Now let’s take a look at the DQ
α
types of the three principals in this crime. OS is a 1.1,1.2; NB a
1.1,1.1; and RG, a 1.3,4. Because none of them possesses a 2 or 3 allele, none is excluded on that basis
from having contributed to the samples.
Both these samples were also analyzed
using the D1S80 system. The advantage of D1S80
is that the interpretation of alleles present is
straightforward — there are no hidden alleles.
The disadvantage is that two D1S80 alleles, 18
and 24, are quite common in the population. The
D1S80 results from both swatches showed bands
at 18, 24, and 25. (One of them is shown below
as CS.) This is, again, clearly indicative of a mix-
ture. Can genotypes be assigned? It depends on
what can be assumed. If it
can be assumed that
there are only two donors, then clearly the 24
and 25 alleles are present as a genotype based on
the similar intensities of the bands as compared
with the 18 band. If two or more donors are
assumed, then the alleles cannot be paired into
genotypes with confidence. No information is
gleaned from either the DQ
α
or D1S80 results
that supports one assumption over the other.
8127/frame/ch10 Page 267 Friday, July 21, 2000 11:38 AM
268
Principles and Practice of Criminalistics
profile indicates a strong link. The probability of the evidence given alternate
sources can be expressed as a likelihood ratio. For
some types of evidence, it
can be difficult to obtain a reliable estimate of the frequency of the charac-
teristic or characteristics exhibited by the evidence. Our confidence in the
frequency estimate (or lack of it) must also be revealed as a limitation of the
test, and factored into the inference of common source.
Do'stlaringiz bilan baham: 
