What Psychological Research Has Told Us about the Accuracy of Eyewitness Testimony

539 Words2 Pages

What Psychological Research Has Told Us about the Accuracy of Eyewitness Testimony

L and P = Loftus and Palmer

Pps = Participants

EWT = Eyewitness testimony

Despite the considerable importance juror’s place on EWT,

psychological research has shown that EWT tends to be unreliable. This

unreliability can be explained in terms of the reconstructive nature

of memory (schema theory). Introduced by Bartlett (1932),

reconstructive memory refers to the extent to which memory is

distorted or otherwise modified (reconstructed) by experience. In

practice this means that instead of storing an exact replica of the

episode they are witnessing, eyewitnesses may combine the initial

stimulus with elements of their existing knowledge and experience (or

schema) to form a reconstructed memory resulting in an inaccurate EWT

account. Reconstruction is not the only source of distortion in EWT as

the language used in leading questions and post-event information may

further distort reconstructive memory and so lead to memory blending

and confabulation. Research that has provided us with an insight into

the effects of language and leading questions on the accuracy of EWT

includes that of Loftus and her colleagues. For example, L & P (1974)

found that changing the wording of one critical question (“About how

fast were the cars going when they (hit/smashed/collided/

bumped/contacted – the five conditions) each other?”) posed to

eyewitnesses of a car accident had a significant effect on the

estimated speed of the vehicles. L and P found that estimated speed

was influenced by the verb used (for example, the mean estimate when

“smashed” was used was 41mph, versus 34mph when “hit” was used). The

verb implied information about the speed, which systematically

affected the pps’ memory of the accident. Further research on the

effect of language (Loftus & Zanni, 1975) found that the definite

article “the” was more leading than the indefinite article “a” as only

Open Document