Factors affecting eyewitnesses (2)

System variable = factors directly under the control of the criminal justice system

Estimator variable = factors outside of the control of the criminal justice system.

Witness and Perpetrator factors:

Witness factors - race, age, gender, confidence, intelligence, attitudes / expectations, quality of description, attention.
Perpetrator factors - race, age, gender, disguise, distinctiveness.

45% of judges surveyed believed it would be more difficult to later identify someone who was wearing a hat at the scene of a crime.

Video tape of a store robbery. The robber was less likely to be identified when wearing a hat (.27 vs .45)

Actors in different conditions - no disguise, hat, sunglasses or both. TP and TA lineups. Correct identification was highest with no disguise, and lowest with both disguises, no significant difference between hat and sunglasses. In TA lineups fewer correct rejections were made with the sunglasses on, although there was no difference between hat or no hat and single disguise or both disguises. This was replicated with a stocking, finding that recognition was worse when 2/3-whole face was covered.

Disguise disrupts global facial configurations.

A disguise means there's less information for encoding, therefore a less complete memory for the perpetrator. This may influence the perception of how difficult the identification task will be. The encoding specifity hypothesis means the disguise may influence attention.

Black and Caucasian faces in a recognition test. Participants are more likely to correctly recognise own-ethnicity faces.

Meta-analysis of 39 studies - mistaken identification was greater in other-ethnicity faces than with same-ethnicity faces.

90% of expert eyewitnesses are willing to testify in court regarding the presence of cross race effect.

Expertise hypothesis - we are more experienced at distinguishing amongst same-race faces. Those with more cross-race experiences show smaller CRE.

CRE is eliminated by asking white participants to closely attend to facial features distinguishing black faces.

Social-cognitive hypothesis - deep processing of faces only occurs for individuals with whom we identify (in-group). White participants viewed white and black faces in wealthy and impoverish contexts (1 in-group, 3 out-group).Only in-group faces were well recognised.

We have more experience telling own-race faces apart (Malpass & Kravitz, 1969). There is a positive correlation of other-race contact and degree of own race bias. Other research has found no relationship (Brigham & Barkowitz, 1978; Malpass & Kravitz, 1969) or mixed results (Chiroro & Valentine, 1995).

An individual's interracial attitudes affects their face recognition ability (Brigham & Barkowitz, 1978) Studies have found no relationship between attitude and CRE (Slone, Brigham & Meissner, 2000). Racial attitudes may play a more indirect role in the CRE, link between racial attitudes and the amount of interracial contact (Meissner & Brigham, 2000).

18-25 and 35-55 year olds shown 4 videos (23 vs 51 year old target) with delay of 1 day or 1 week in a target present lineup. An own age bias and effect of delay was observed.

18-32 and 55-85 year olds shown videos of 18 and 54 year old target in TP lineup. There was no evidence of own age bias. Young adults were more likely to correct with the older lineup, older adults showed no effect of age on lineup accuracy.

Situational factors of estimator variables include: stress and arousal, weapon presence, view of the perpetrator, influence of alcohol, own race identification, own age identification and own gender identification.

Alcohol impairs perception by narrowing the focus on central aspects of an event.

Intoxicated eyewitnesses are less likely to remember peripheral details compared to central details.

In TP lineups there was no effect on correct identifications, although in TA lineups there were more false alarms.

In a hypothetical sexual assault scenario, participants had either no, medium or high alcohol and a recognition test. Peripheral details were remembered less accurately than central details. There was no effect of alcohol on memory accuracy.

TP and TA lineups showed that as distance increases and illumination decreases, recognition performance decreased. Rule of thumb of no more than 15m.

Participants viewed 8 targets at one of 8 varying distances (5, 10, 15, 20, 25, 30, 35, 40 yards). Increased distance associated with fewer correct recognitions and more false alarms.

Errors in change detection can be an explanation for unconscious transference. 60% of participants chose a previously viewed bystander in a target absent lineup.

Change blindness shown in video taped crime with two innocent bystanders in a shop. 59.6% of participants failed to notice the change between the innocent bystander and the perpetrator of a crime.

Co-witness effect: a description provided by a group of witness is more complete than one by an individual.

Conformity effect: witnesses viewed events which had key differences and were then asked to discuss it. 71% of participants incorporated erroneous details that were provided by other witnesses. This shows a contamination of memory.