Mathematics and the Court
System?
In the field of mathematics and statistics many people feel
that the only ethical issues that apply are those dealing with plagiarism,
fudging calculations, and/or ethical issues associated with teaching
mathematics in k-12 schools. However,
there are many other areas in the field where ethical issues come into
play. For example, in the field of
statistics there is a huge debate that deals with who should interpret
statistics associated with court cases, specifically DNA profiling. The question is should the courts be the ones
to interpret the data that the statistician has collected/formulated or should
the statistician provide a final interpretation/conclusion (i.e. which person
is more ethically correct to choose/pick).
To an outsider or a person in the field of science and
mathematics it seems that the answer to this question is fairly obvious; the
statistician would be the most qualified person to interpret the data and offer
a final interpretation/conclusion.
However, many people are skeptical of statistics and fear that the
results of a particular set of data can be analyzed in such ways that, even
though accurate, provide a false picture of what the data actually
represents. This can be seen in the
classic saying that “Statistics are like bikinis. What they reveal is suggestive, but
what they conceal is vital.”
Personally, as a person in the field of mathematics, I feel
that it would be unethical/immoral to have a court interpret the statistics
because in my eyes they are not the most qualified person to do so and
therefore would open the door for more mistakes and misinterpretations of the
data that could severely affect the person on trial in a negative way.
Below is an excerpt from an article in Mathematical
Association of America (MAA) that deals with the state of California
“On
August 16, 2006, the California Supreme Court made it official: in certain
legal cases that hinge on statistical calculations, it is not the business of
professional statisticians to decide how to evaluate the statistical data and
to judge what method is most suited to analyze that data. From now on, in California
Now that we have a general idea about the ethical debate of
deciding who should analyze/interpret the data we can look into some of the
statistics that go into suspect identification and conviction based on DNA profiling
(the specific issue at hand).
As a standard, the FBI looks at 13 regions of the DNA strand
to match completely in order to reduce the error of false identification. The idea behind this is that since the
probability that anyone would match a specific DNA strand at any one point is
(1/10), that by using the product rule for multiplying probabilities, taking 13
regions would give us a chance of matching a given DNA sample at random in a
population are about 1 in ten trillion (i.e. 1 / 10,000,000,000,000). This figure is known as the random match
probability (RMP).
This number leaves little room for error when profiling DNA
samples, but to attain this number a qualified professional needs to understand
which formulas to apply and how to interpret the results based off of the
chosen formulas. Something I feel a
court is highly unqualified to do.
Overall, it seems that the courts want to uphold the famous
saying “guilty beyond a reasonable doubt”, and by trying to do this they are
attempting to eliminate any possible bias that may come from the
evidence/statistics they consider. This
seems reasonable, but I feel that a person needs to weigh the consequences of
each action (i.e. consequences associated with letting courts select formulas
and interpret statistics vs. consequences of letting a statistician do the same
jobs). I feel that the probability of
wrongful conviction greatly increases when untrained professionals handle the
statistics and feel that training statisticians on the ethics associated with
the courts would suffice to make up for any biases the courts fear.
If you’re interested in the California
http://www.maa.org/devlin/devlin_09_06.html
