Factors that affect Eye Witness Testimony

Emotion

Research shows that emotion can either improve or reduce recall. The Yerkes-Dodson law states that recall improves with arousal up to an 'optimum point'. Moderately frightening produces the best recall and highly fightening the worst.
Deffenbacher et al (2004) conducted a meta-analysis and found that high stress events had a negative impact on testimony accuracy.
Yuille and Cutshall (1986) however found that witnesses who were most distressed by their experience were most accurate in their recall, however their sample was small so findings can't be generalised, their participants witnessed bank robberies, thus findings can't be generalised to more personal crimes.
Macleod et al (1986) studied 379 reports of eyewitnesses to assault compared to witnesses of crimes where no injury was caused and found no difference in recall.
These studies highlight findings in this area are not consistent.

1 of 5

Clifford and Hollin (1981) showed that percieved seriousness of a crime affects eye witness testimony. They showed participants one of two films, violent and non-violent interaction between a man and a woman. Memories of those shown the violent film were less accurate and less complete.
Loftus and Burns showed participants a bank robbery, one with a violent ending and the other without. They found those shown the violent version gave a less accurate and less complete account of events. Both these studies however were conducted in lab settings, may have been affected by demand characteristics, or people not trying very hard as it wasn't real.
Christianson et al (1993) questionned 110 witness and found those who were victims had better accuracy compared to onlookers. It was a natural experiement so has high ecological validity however lacks control over extraneous variables.

2 of 5

Loftus, Loftus and Messo (1987) claim the presence of a weapon interferes with a witness's recall of an incident and assailant. During a crime people fixate on the weapon.
Loftus et al used two conditions, in both participants heard a discussion in an adjoining room.In one a man emerged holding a pen, with grease on his hands, and in the other a man emerged holding a letter opener covered in blood.
When asked to identify the man participants in condition one were 49% accurate and just 33% accurate in condition two.
Johnson and Scott (1978) using a field experiement split people into two conditions, in one they saw a man carrying a knife covered in blood and in the second they saw a man carrying a pen covered in grease. Recall was worst in the first group. This study has higher ecological validity.

3 of 5

Carmichael demonstrated that our memories are affected by language and schemas. Groups were given a set of pictures, and then a set of labels. They were then asked to produce the pictures. Results showed the label affected the picture produced.
Loftus and Palmer found that the severity of the verb used affected speed estimation (smashed - 40.8 and contacted 31.8). When asked later 'did you see the broken glass?' 14% in the 'hit' condition said yes and 32% said yes in the 'smashed' condition. All participants were however students so the findings cannot be generalised.
Moreover Yuielle and Cutshall (1986) interviewed 13 witnesses of armed robbery and found that leading questions did not impact their testimonies. The sample was both small and from Canada so findings cannot be generalised.

4 of 5

People often misidentify offenders as we are poor at recognising faces. Buckhout (1974) staged a purse theft and conducted two line ups to test the recall of 52 witnesses, only 7 correctly identified the culprit on both occasions.
Buckhout and Regan (1988) highlighted that we are especially poor at recognising faces from other cultures; referred to as cross-race effect.
Ellis et al (1979) found face outline and hair style were important when identifying an unfamiliar face but features such as eyes were important for recognising familiar faces.
Brown et al (1977) highlighted that although we sometimes recognise unfamiliar faces we are not always able to place where we have seen them which leads to wrongful identification.

5 of 5