by Cindy Rennie
Al D’Silva
2009-01-23
On January 23, 2009, Dr. Itiel Dror, Senior Lecturer in the
Department of Psychology at the University of Southampton (UK) gave
a lecture at the
University of
Toronto
entitled “Why forensic examinations are inherently biased, and what
can be done about it.”
Dr. Dror has spent many years researching the effect of
psychological and contextual factors on decision-making in forensic
examinations. He has chosen to focus on the
fingerprint field because it has been around for over a hundred
years, it is well understood worldwide, and it carries a lot of
weight in court.
The complete list of his experiments (and their findings) may
be found at http://users.ecs.soton.ac.uk/id
Dr. Dror began his lecture by listing what he calls “the
ingredients of decision-making”:
1.
Data – decisions are
based on the information available at the time.
2.
Parameters – decisions
are affected by such factors as the number of alternatives available
to the decision-maker, the complexity of the decision, and the
amount of time available to them in which to make the decision.
3.
The decision-maker
themselves – decisions will be influenced by such factors as the
age, gender, education, and level of experience of each
decision-maker.
One of the questions that his research asks is, “Is
decision-making consistent?” or would a decision change if the
circumstances were changed?
Dr. Dror believes that there are two sets of
circumstances that affect a person’s decisions:
internal and external. Internal circumstances
include such things as each person’s point
of view, level of expertise, education and values.
External circumstances are the resources available to
each person and the circumstances surrounding the decision (e.g., is
there a “right” or “wrong” answer, and what are the consequences?)
Dr. Dror’s experiments try to determine if the
decision of a fingerprint examiner is based on the actual data and
evidence available, or if it can be affected by context.
He made reference to three of his studies of fingerprint
examiners (all of which are available on the above-noted website)
“When Emotions Get The Better Of Us: The Effect of Contextual
Top-down Processing on Matching Fingerprints”; “Contextual
Information Renders Experts Vulnerable to Making Erroneous
Identifications”; and “Why Experts Make Errors”. (He mentioned that
there were more studies on the website.)
Here is an example of the kind of experiments that he is
performing:
The reader is most likely aware of the false identification
to Brandon Mayfield that was made in the
Madrid bombing case.
Dr. Dror chose five fingerprint examiners from a global list
of volunteers, each of whom had approximately seventeen years’
experience in examining fingerprints, and none of whom had seen the
Mayfield print. The volunteers did not know when
or how they would be tested. The fingerprints were presented to them
in the normal course of their duties.
Dr. Dror arranged to have a latent fingerprint
retrieved from the files of each examiner - a latent fingerprint
that had been identified by that examiner five years before. That
same latent was presented to the examiner again along with the
‘known’ fingerprints of the suspect that they had identified, only
this time the examiner was told that the latent and ‘known’
fingerprints were the ones involved in the false identification from
the
Madrid bombing case.
The fingerprint experts were asked to decide whether there
was sufficient information available in the pair of prints to make a
definite and sound decision, and, if so, whether they found the two
prints to be a ‘match’ or ‘non-match’. They
were instructed to ignore the context and background information,
and focus only on the actual print in their evaluation and
decision-making.
Of the five examiners, only one stuck to his original
decision and declared the two prints an ‘identification’.
The other four changed their decision – three said that the
prints were a definite ‘non-ident.’, and one said that there was
insufficient quality of detail to individualize.
(Remember, each examiner was looking at a set of prints that had
come from their own caseload and that they had already declared as
an identification in the past!).
Dr. Dror states that this is an example of how the decisions
of fingerprint examiners are vulnerable to irrelevant and misleading
contextual influences. He says that it is impossible to ignore
context. Once an examiner is exposed to the
context surrounding an examination, (such as the type of crime
involved, or the fact that the investigators have other evidence
that implicates this individual) they cannot put that context out of
their mind – the examiner cannot ‘unlearn’ something that may affect
their examination of the fingerprint.
His new study examines the phenomenon of errors in
fingerprint examinations.
He suggests that inconsistent identification decisions
may be caused by ‘cognitive flaws’, which place limitations on the
ability of the mind to conduct objective and independent processing
and evaluation of information.
The mind is not a camera,
Dr. Dror explains. Each person’s mind perceives data differently.
Each person’s perception is influenced by such individual
psychological issues as ‘attention span’, ‘confirmation bias’,
‘wishful thinking’, ‘escalation of commitment’ and ‘self-fulfilling
prophesies’. Dr. Dror is trying to determine if
these factors lead to errors, and, if they do, are they
‘practitioner’ errors, or are they caused by a flaw in the process
itself.
Dr. Dror showed some examples of how the mind can be
manipulated by outside factors. He told the
audience that he was going to show a video of two teams playing
basketball. Each team had four people on it. One
team was dressed in white, one was dressed in black.
We were told to count the number of times that the members of
the white team passed the ball without bouncing it first.
After the video was played, members of
the audience had a wide variety of answers. Some
saw the white team pass the ball as many as seventeen times, others
saw them pass the ball far fewer times. What
nobody saw was the six-foot man in a gorilla suit who strode into
the middle of the picture, waved his arms around, and strode off!
We didn’t believe that there HAD been a six-foot
gorilla in the video – until the video was played a second time.
There he was, big as life. We had been concentrating so much
on watching the white players and the ball, that we did not see the
gorilla.
Dr. Dror explained that this is similar to an
investigator (or fingerprint examiner) developing “tunnel vision”
when working on a case. If you concentrate on one
area of the fingerprint, you run the risk of missing other
information that you need to make an informed decision.
In another example, Dr. Dror put a phrase on the
screen and asked the audience to count the number of times that the
letter “f” appeared. Most of us counted five. It should have been
eight. The word “of” appeared three times, and nobody had counted
the “f”s in them. Dr. Dror explained that the
mind uses a kind of short-hand, where it recognizes and ignores some
commonly seen words, such as “of”.
He pointed out that these types of errors in
perception and bias by context seem to have less influence on an
expert when dealing with excellent prints of high quality.
The effects become more pronounced when dealing with poor
quality prints that require more analysis.
The short-term solution to the problems that Dr. Dror
is studying lie in the selection and training of fingerprint
examiners and in the procedures and protocols that govern each
workplace.
Dr. Dror also states that more research needs to be
done into these topics.
Some of that research is being done by the
National Academy of Sciences (NAS) (which is expected to publish the
results of their latest research into the field of forensics
shortly) and the National Institute of Studies in Technology (NIST).