5. Eye witness testimony

what is eye witness testimony?

account/evidence provided by people who witnessed an event, such as a crime, reported from memory

what are leading questions and how do they disrupt recall?

information that leads you to giving a particular response, as opposed to an accurate response. A subtle change in the wording of a question can have a dramatic effect on recall. This is because our memories are not precise copies of what happened- we only maintain fragments of info- and we fill in the rest (episodic memory). Reconstruction of events leads to inaccurate recall- can be compounded if we recieve info after the event, which distorts our original memory.

research into misleading info study 1

• Loftus and Palmer

• ps (45 uni students) showed short films of automobile accidents and then answered qs about them in a questionnaire (avoid demand characteristics).

• Critical Q- ‘How fast were the cars going when they ____ into each other?’

• smashed- 40.5 mph (connotations of speed + severity)

• bumped- 38.1 mph

• contacted 31.8 mph

• memory of incident changed due to leading q- qs subsequent to an event distort reconstruction- affects way a memory is recalled

research into misleading info study 2

• Loftus and Palmer

• ps (150 students)- 1 group asked ‘smashed’, one ‘hit’, one no q- control group. Returned a week later and asked if there was any broken glass

• smashed- 16 yes, 34 no

• hit- 7 yes, 45 no

• control- 6 yes, 44 no

• memory of incident changed due to leading qs. reconstruction of event (adding in things to recollection)

misleading info evaluation- validity

p- issue with validity of Loftus and Palmer’s research

e- witnesses of a real life car crash likely to experience a greater response than those watching the film- affect memory of witnesses. also in real life may be focussing on different things- attention not only on the crash- lacks ecological validity

e- ps may have worked out aims of study- showed demand characteristics, internal validity. Also, only carried out on uni students- lack of driving experience- lacks generalisability.

l- findings may tell us little about EWT in real life

misleading info evaluation- reliability

p- high reliability- many lab studies into leading qs have found similar things

e- the findings are consistent and the studies are repeatable- as take place in a highly controlled environment- high internal validity- can checl reliability

e- HOWEVER, one experiment by Loftus found that leading qs did not change the accuracy of EWT when the info was OBVIOUSLY INCORRECT.- Ps showed slides of a theft of a large purse from a handbag, 98% rememberd correct colour. Then read an account of the theft- allegedly written by a professor- which said the purse was brown (it was actually red). Only 2 ps changed their answer to brown on a subsequent test.

l- cannot reliably conclude that all misleading info will have the same impact in all circumstances- e.g. if obviously different

misleading info evaluation- application

p- important practical applications

e- used to train professionals e.g. police should not use suggestive words/questions (leading qs) to avoid misleading people. Also, psychologists have used research in this area to develop the cognitive interview- to improve accuracy of EWT

e- positive economic implications- conserve police money, resources and time- reduces liklihood of police obtainig unaccurate info

l- importance and value of this research

post-event discussion

When co-witnessed discuss the event (which is likely) it can lead to further distortion- come into contact with post event misinformation.

could be due to ISI- ps may genuinely believe that they misremebered something and think the co-witnessed know more than them so adopt their recollection of events in their own reconstruction

evidence into post-event discussion

• Gabbert et al

• ps (60ps students from university of Aberdeen and 60 older adults recruited from the local community)

• 2 conditions-individual tested (control group), in pairs (co-witness group)

• in co-witness group- 1 watched clip where girl stole money/2- watched perspective where you don’t see girl stealing money

• after watching their video clips the co-witness discuess the crime together and all ps completed a questionnaire testing their memory.

• co-witness group- 71% recalled info they had not actually seen vs 0% control

• co-witness group- 60% believed girl was guilty despite not seeing the crime take place

• highlights issue of PED and its powerful effect on the accuracy of EWT

evaluation of PED validity

p- lacks mundane realism- controlled nature of research

e- an eyewitness to a real crime who is then involved in PED may not act the same same as ps under lab conditions. e.g. may have been effect of conformity in this research as they knew it was a study

e- outcome of ps answers had no social impact, no one would suffer as a result of their testimony (i.e. prison)- make them more likely to conform and recall info they didn’t see- no consequences for doing so

l- lack of validity In lab research- must be cautious when generalisin to real life

evaluation of PED- reliability

p- consistency of findings

e-many research studies have found similar findings. e.g. Gabbert et al found that ps recieving misinformation after an event were less accurate at recalling an event than controls

e- especialy if this info came from a social source compared to a non- social source (written)

l- research has high reliability - confidently suggest that EWT can be distorted by PED

evaluation of PED applications

p- practical applications

e- e.g. police officers should advise witnesses not to discuss the case with other co-wtinesses asap to prevent testimony being distorted by others

e- however, in the aftermath of an incident it is natural for eye-witnesses to discuss what they had just seen- so police officers must take that into consideration when later questioning them.

l- value of these research findings

Impact of anxiety on EWT

eye-witnesses are often very stressed and anxious when witnessing a crime, especially if they are victims. This is because of the danger they percieve to be in. Anxiety can be a source of distortion for EWT

Yerkes-Dodson effect

when arousal is very high or very low it has a negative impact on performance but moderate levels are beneficial (physiological arousal= pounding heart, rapid shallow breathing etc).

Yerkes-Dodson effect EWT

in cases such as a violent crime, witnesses are likely to be in a state of high arousal- so performance in recalling details is negatively impacted

Weapon-focus effect

Loftus- anxiety is more problematic for recall when crime involves a weapon. Research suggests that the anxiety of seeing the weapon focuses all your attention on the central details (the weapon), less able to recall peripheral details (what else is going on, perpetrator’s appearance)

weapon focus effect research- 1979

• Loftus

• ps either heard a hostile and aggressive arguement and saw someone emerge holding a bloody letter opener or ps hear a harmless conversation and see someone emerge holding a greasy pen

• letter opener- 33% correctly identified culprit

• pen- 49% correctly identified culprit

• anxiety caused by seeing the letter opener caused the WFE- which led to ps not paying attention to other info- hence poor recall of culprit

evaluation of role of anxiety in EWT- internal validity

p-may lack internal validity- one criticism of WFE is that it may not be caused by anxiety

e- critics suggest it is testing SURPRISE not ANXIETY

e- Pickel- showed videos of a thief either entering a hairdressers with gun, scissors, wallet or raw chicken, then asked ps to correctly identify the thief. Ps accuracy was lowest for the chicken, then the gun- perhaps due to the fact that they were surprised by the chicken as it is so highly unusual

l-suggests WFE is due to unusualness/surprise rather than anxiety/threat- other studies suggest nothing specific about role of anxiety

evaluation of role of anxiety in EWT- reliability, ecological validity and generalisability

p- differences in the results between research conducted in a lab environemnt and research on real-life eye-witnesses- lack of consistency

e-Yullie and Cutshall - 13 people who had witnessed a real life shooting in Canada were interviewed 5 months after the event. ps were asked to recall info about the event and also report whether there anxiety levels at the time were high or low

e- high anxiety- 88% accuracy, low anxiety- 75% accuracy. Maybe fight or flight was activated by physiological arousal - increases alertness and improves memory as we become more aware

l- contrast to Loftus’ lab findings- can’t be generalised to real life

evaluation of anxiety- application

p- important real world applications

e- role of anxiety in EWT- police proceed with extreme caution when pursuing lines of inquiry

e- criminal investigations rarely rely solely on EWT and also use other evidence (e.g. CCTV or DNA) to corroborate testimony before convicting someone

l- research has positive benefits for criminal justice system- reduce chances of convicting the wrong person

reasons why we forget

• state & context dependent forgetting

• proactive & retroactive interference

• misleading info- incorrect memory reconstruction

• high/low arousal- weapon focus effect

how do police traditionally interview eye witnesses?

• frequently interupt them

• ‘focus on relevant details

• use q+a format

`key ideas behind the development of the CI

• memory is complex- many types

• retrieval cues are important and should match- Tulving’s ESP- to avoid inaccurate recall

• misleading info can distort recall

who created the CI

Geiselman et al

techniques used in the CI

1. recreate the context of the original incident

2. report every detail

3. recall the event in reverse order

4. change perspectives

recreate the context of the original incident

witnesses asked to pictures the surroundings of the crime and visualise the settings (e.g. weather and how they were feeling)- IMPROVE accuracy of recall as context and state cues are the same- Tulving’s ESP. Different to police interviews- just asked to free recall and answer qs not recreate context

report every detail

tell witnesses not to withhold any information, even if they think its irrelevant- should be encouraged to tell the whole story(leaving nothing out) interviewer must be patient. This will help the p remember more by triggering cues for retrieval of a large amount of accurate info from memory. Avoids leadings qs. Different to police interviews in which the repetition technique is used- repeat the questions needed and frequently interrupt witness.

recall the event in the reverse order

get the witness to try an tell the story in the revere order. This stops witnesses from recreating the event in relation to expectations/stereotypes about what normally happens in crimes. Recalling the crime in a variety of different contexts is key to cueing retrieval of info. Different to police interviews in which they usually ask for free recall- leads to chronological retelling of events

change perspectives

ask witnesses to recall info from a different physical perspective (what might they have seen from across the street? what did other witnesses see?) Disrupts personal expectations- prevent inaccurate reconstructions. Triggers cues that lead to accurate recall. Standard police interviews only focus on witness’ perspective

evaluation of CI- research evidence and reliability

p-use of ci is supported by research

e- Geiselman assessed whether CI is more effective than standard interviews when interviewing eye witnesses. 89 students watched police training videos. 48 hrs later the students were interviewed individually by American Law enforcemnt officers either using standard police interview or cognitive interview. Led to more accurate info being recalled (41.5 vs 29.4) However, amount of incorrect info was also higher (7.3 vs 6.1).

e- The findings from both Kohnken and Geiselman’s research studies demonstrate consistent results and clearly show that CI improves accuracy of EWT

l- cognitive interview does lead to higher level of

evaluation of CI- meta-analysis/research evidence

p- supporting evidence

e- Kohnken et al carried out a meta-analysis of 53 studies investigating the enhanced cognitive interview. On average, 34% increase in amount of correct info generated compared with standard police interview.

e- however, these studies are lab experiments with volunteer ps- must be cautious when drawing conclusions about real life

l- supports use of ci over standard police interview

evaluation of CI- economic impact

p- limitation- tends to take longer and requires more training than standard interviews

e- problem because it has negative econmic implications- more expensive to carry out, on the other hand - positive economic implications - don’t arrest/imprison wrong people

e- Therefore a compromise is needed- Milne and Bull found that a combination of report everything and context reinstatement techniques produce the most effective results

l- while all 4 elements may not be practical, research into CI has found that only using two techniques is extremely effective