Wells, G. L. & Bradfield,
A. L. (1999). Distortions in eyewitnesses' recollections: Can the postidentification
feedback effect be moderated? Psychological Science, 10, 138-144.
(A
draft of this paper is available online)
Background
Eyewitness confidence is the biggest
determinant of whether or not a witness will be believed by members of
the jury.
Unfortunately confidence isn't a great
predictor of accuracy. One reason for that is because witness confidence
can be influenced by things in addition to how strong their memories are.
One such factor is post-identification
feedback, telling a witness that they are right or wrong in their identification
(Wells
& Bradfield, 1998).
Postidentification feedback not only
increases confidence, it makes witnesses think they were always
confident and also makes them believe the witnessing conditions (e.g. how
good of a view they had) were better than they really were!
This is all the more troubling because
witnesses do not think that the feedback is having any effect on them.
It may be possible to "inoculate" witnesses
against the PIF effect by having them think about things like viewing conditions,
confidence, etc. before receiving any feedback.
Method
Subjects saw a robbery video and then
received a target absent photo lineup. The situation was set up to imply
that the gunman was in the lineup and everyone made a false identification.
Subjects were randomly placed into
one of 5 experimental conditions based on whether they got confirming feed
back and whether they were asked to think about the witnessing conditions
No Thought
about Witnessing Conditions
Thought
About Witnessing Conditions Prior to Feedback
Thought
About Witnessing Conditions After Feedback
Confirming
Feedback
No Feedback
XXXXXXXXXXXXXX
The effect of PIF on confidence was
reduced by having subjects think about the witnessing conditions before
receiving feedback. Unfortunately this seems to have come at a cost as
prior thought (and subsequent thought) appear by themselves to increase
witness confidence in their false identifications. Subsequent thought did
not eliminate the effect.
For most of the other measures (view,
details of face, amount of attention, basis for ID, ease of ID, speed of
ID, willingness to testify, ability to ID strangers) the effect of PIF
was reduced by having witnesses think about the witnessing conditions before
receiving the feedback and in some cases it appeared to be eliminated altogether
(see figure 2)
For three measures (view, ease of identification,
speed of identification), the PIF effect was reduced by subsequent feedback.
Important Findings
PIF had big effects on witness confidence
and their estimate of ease of identification, their overall memory ability,
how good of a view they had and so on. This is not a good thing!
Thinking about witnessing conditions
before receiving PIF (and to a certain extent afterwards) reduced the effect.
This is a good thing!
Thinking about witnessing conditions
before receiving PIF did not entirely eliminate the effect. This is not
a good thing!
Possible to explain results in terms
of self-perception theory (an interesting theory for those of us interested
in consciousness). Self-perception theory says we don't necessarily have
privileged access to our own thoughts, perceptions and so on (not to mention
so forth). Instead, at least partly, we infer our own mental states
in much the same way other people infer our mental states, by watching
our behavior. If I got the suspect, then I must have had a good
view.
Prior thought by itself leads to confidence
inflation too. This, you guessed it, is not a good thing. However, the
effect seems limited to confidence, not the other measures. And that is
a good thing!!!!!
Perhaps the best thing would
be for police not to provide PIF at all. However, that may not be possible
at all in principle. Afterall, at some point witnesses will learn that
the person they fingered is being charged with a crime. It may be important
to consider if feedback will continue to have an effect if it is delayed.
Additionally, this fact underlies the importance of obtaining a clear statement
of confidence from the witness before they receive any feedback.
Anyway, that's the end of the summary
and that....is a good thing!!! :)