Cognitive Psychology
Eye witness testimony: Loftus et al (1978)
Abstract:
1242 Pps, in 5 experiments plus a pilot study saw slides of a car
accident. Loftus et al aimed to see how information supplied after an
event influences the witnesses' memory of the event. Pps exposed to
either consistent or misleading information (IV) after viewing slides and
recall was tested (DV). Also the time when the Pps were given the
information was varied, either immediately after viewing or just before
questioning (IV) and recall was tested (DV). It was found that
consistent or misleading information was integrated into the witnesses
memory and this effect was strongest when the information was
presented just before the questioning.
Background information:
Kant (1781) first suggested that humans tend to merge different
experiences to form new concepts and ideas. This tendency causes
particular concern in the field of EWT. Loftus began investigating this
tendency objectively in 1974.
Method:
This passage describes just one of the five experiments undertaken by
Loftus et al (1978).
Design: Controlled experiment using an independent measures
design. Experimental group presented with misleading information (IV)
after event and recall tested (DV). Control group presented with
consistent information (IV) after event and recall tested (DV).
Participants: 195 students from Washington university.
Materials: series of 30 coloured slides, questionnaire of 20 questions of
which provided either the misleading or consistent information, an
unrelated short story with recall questions which acted as a filler task,
15 pairs of slides (one old one new) shown on two slide projectors.
Procedure:
1. Participants shown the series of 30 slides of a small red Datsun car.
The critical slide showed the car stopped at either a stop sign or give
way sign.
2. Participants given a questionnaire of 20 questions about the slides
seen. Half the Pps had a question which asked “did another car pass
the Datsun at the stop sign”, the other half had a question which said
“did another car pass the Datsun at the give way sign”. So, half the
Pps had a question which was consistent with the slide they had seen
and for the other half, it was misleading, e.g. it asked a question about
a sign they had not seen.
3. Next the Pps were asked to complete the filler activity of reading the
short story and answering questions.
4. Finally they were tested used a forced recognition task where they
were shown 15 pairs of slides. In each pair, one was a slide they had
seen and one was not. The Pps were asked which they had seen. The
critical pair in this activity showed the Datsun at either the stop or give
way sign. The experimenters were measuring whether the Pps
identified the real slide they had seen or the slide which they had been
led to believe they had seen in the questionnaire.
Results: Pps correctly identified the correct slide when given
consistent information in the questionnaire 75% of the time and only
41 % of the time when they were given misleading information. When
the data was analysed statistically, those given consistent information
performed significantly better than chance and those given misleading
information performed significantly worse than chance.
Discussion: Misleading information presented after the initial
encoding of information caused less accurate retrieval. The Loftus
experiments support the view that memory involves interpretation and
reconstruction of the past. We try to fit memories into our existing
schemata to make them more logical, coherent and sensible. We draw
inferences and make deductions about what might have happened or
in our opinion should have happened. This describes memory as an
active attempt to understand our experiences.
Conflicting evidence also gathered by Loftus shows that memory is not
always so fallible, when it is important details of the event which are
presented either consistently or in a misleading way, Pps are mush
harder to fool. The effect seems only to be true for peripheral
information. When Pps asked about colour of a purse which they see
being stolen, only two were mislead by being told the purse was a
different colour. Concentrating on the mistakes we make gives a rather
one sided view of memory.
Conclusion: The numerous experiments carried out on EWT by Loftus
and her many colleagues suggest that memory is recontructive and
can be shaped by misleading information and leading questions. The
Devlin report (1976) recommended that trial judges instruct juries not
to convict on EWT evidence alone except in exceptional circumstances
however Brown (1986) says a person can be convicted under criminal
law on EWT alone so there must be many miscarriages of justice.
Jurors seem to find this kind of evidence very persuasive even if
psychologists, judges and defence lawyers do not.
Criticisms of EWT research:
Ecological validity
Concentration on fallibility of memory and not accuracies of
important details: A study by Yuille and Cutshall (1986)
showed that EWT in a real life situation can be highly reliable.
(see additional notes)
