Pozzulo,
J. D., & Warren, K. L. (2003). Descriptions and identifications of
strangers by youth and adult
eyewitnesses. Journal of Applied Psychology, 88, 315-323.
There
were two goals of this article: to examine the amount and accuracy of
descriptions and to determine if there is a relationship between the length of
the descriptions and identification accuracy for both youths and adults.
Based
on prior research, it was hypothesized that youths would provide fewer
descriptors than the adults but that there would not be a relationship between
number of descriptors and accuracy, and that they would be less accurate at
giving body descriptors and targets’ ages,
Experiment
1
Methods
Youths in grades 5 – 8 and undergraduate psychology
students were shown a “street proofing” video of Mike that described ways to be
safe on the street. After the video was
shown, they were asked to describe Mike, completed a filler task, and then
attempted identification of Mike in either a target-present or target-absent
lineup.
Youths reported fewer descriptors and were less accurate
than adults.
Not
only did youths report fewer descriptors for the categories of interior face descriptors
and body descriptors, but they were less accurate than the adults.
Youths
also reported fewer descriptors in the category of exterior face descriptors.
Adults
reported fewer accessories.
Youths
were also less accurate at reporting the age of the target, and they were less
likely to report the target’s race.
In
the target-present lineup, there was not a significant difference of
identification accuracy across age.
In
the target-absent lineup, youths were less likely to reject the lineup (thus,
they made an identification even though Mike was not there).
The
number of descriptors did not differ between accurate & inaccurate
identifiers, for both youths and adults.
Methods
Instead of watching the videotape of Mike, Carol came into
the classroom and gave a little talk about the study (youths) and recruiting
participants (undergraduates). The rest of the methods were very similar to
Experiment 1.
Results
Like in experiment 1, youths reported fewer descriptors
than adults, however, they were not less accurate. (But, they did report very
few descriptors in this experiment, limiting their opportunity to be incorrect,
according to the authors. Further, they
claim that a live target could reduce the number of descriptors provided due to
increased levels of anxiety or arousal).
Youths were less accurate and reported fewer total
descriptors in the categories of body descriptors, clothing descriptors, and
interior face descriptors. They also
reported fewer exterior face descriptors and accessories.
In the target-present lineup, there was not a difference
of identification accuracy across age.
In the target-absent lineup, there was not a difference in
the numbers of correct rejections as a function of age (they were equally as
likely to say that Carol was not in the lineup.)
In the target-present lineup, more descriptors were
reported for accurate identifications than for inaccurate ones.
In the target-absent lineup, youths gave fewer descriptors
than adults for both correct rejections and false positive identifications.
Discussion
Seeing Carol, a live target, as opposed to Mike, a
videotaped target, led to fewer descriptors being reported by both youths and
adults. Could it be that a live target is
more engaging or distracting? If so, then this is problematic, because crimes
are live. The authors state that we need to better understand the relation
between arousal and recall memory for youths (and probably adults as well).
Youths had difficulty reporting interior facial features
and body descriptors, as compared to the adults. The authors state that this could be a developmental trend where
exterior features are initially more salient or that interior features may
require a richer vocabulary to describe, or it could rely on gestalt
processing, which may occur later in development.
Reported age of the targets needs to be considered in
relation to the age of the witness.
Why? Because adults were accurate at describing Mike, who was close to
their age, but not very accurate at describing Carol, who was more like the age
of their parents.
In conclusion, from a practical perspective, it does not
appear that description length should be used to assess the witness’s ability
to make an accurate identification for either youths or adults.