Eyewitness Testimony

Question 1

Episodic Memory

Witnesses asked to remember what happened

Answer 1

Memory for experiences

Literal record but reconstructive process.

Remembering is not like replaying video, but rewriting a story using schemas.

Question 2

Episodic Memory

Evidence

Answer 2

Memory of story content is affected by the ending

Memory of story content is affected by expectation; changing expectation changes memory of content

Question 3

Episodic Memory

Implication

Answer 3

Memories may largely reconstructed

Reconstructions are fallible

Eyewitnesses can be very fallible

Question 4

Stress and Memory

How does stress affect memory?

Answer 4

Complex interaction

Some stress can be good, too much can be bad.

Question 5

Stress and Memory

Is stress bad for memory?

Answer 5

Children’s memory for events in a hurricane correlates with damage to house; memory / stress relationship consistent with quadratic function.

Question 6

Stress and Memory

Why quadratic?

Answer 6

Yerkes-dodson curve

Moderate stress and arousal benefits performance

Question 7

Stress and Memory

Stress narrows processing

Answer 7

Stress limits attention to central details;

weakens memory for peripheral details.

Question 8

Malleable memory

Suggested response bias or suggested memory bias?

Answer 8

Response to a question may be influenced by phrasing of the question.

“do you get headaches? how occasionally / frequently?”

Is this a change in memory change in response?

Question 9

Malleable memory

Phrasing biases reported memory for details

Answer 9

Estimates of speed are influenced by phrasing?

How fast was the car travelling when it contacted / smashed / hit / bumped / collided into the other car?

Memory for detail is affected by phrasing?
“did you see a broken headlight?” 7% yes
“did you see the broken headlight” 17% yes

Question 10

The misinformation effect

Answer 10

Post-event information may alter memory for event information.

Misinformation
“did another car pass red car at the yield sign?”

No Misinformation
“Did another car pass red car at the junction?”

Test
Which of these slides did you see?

No misinformation - 75%
Misinformation - 41% correct

Question 11

Misinformation Effect

What is it?

Answer 11

Personally witness an event
Read / hear incorrect details about that event afterwards

Incorrect details may infect your memory of the event.

Report what you witnessed

Question 12

Misinformation Effect

Source Memory Misattribution

Answer 12

Attributing information in memory to the wrong source

Question 13

Misinformation Effect

What affects it?

Answer 13

Several factors affect susceptibility, such as age and working memory capacity.

Environmental factors moderate the influence of misinformation, such as time delay and exposure time.

Physiological factors may increase the risk of misinformation, such as alcohol withdrawal, sleep deprivation.

Incorrect belief in compromised memory, such as false alcohol ingestion

Question 14

Differences in memory metrics

Answer 14

Reaction time and confidence

Misled condition - misleading post-event information
Control condition -
neutral post event information

Reaction times and the confidence ratings show that misled subjects are certain of their responses.

Incorrect misled responses are indistinguishable from correct responses.

Question 15

Powerful Misinformation

Sam Stone study

Answer 15

Classic study 
3-6yrs misled about behaviour of class visitor called Sam Stone. 

Given stereotype (pre-event)or suggested (post-event) misinformation or both.

Both types of manipulation result in false reports.

Over 20% ppts receiving both types of manipulation maintained their reports were true even when challenged.

Trained professionals could not distinguish between truth and false reports on video footage of ppt testimony.

Powerful memory error

Question 16

A social contagion

social misinformation

Answer 16

Co-witness recall of items in a picture

Co-witness recall of details of a video event
- 2 ppts watch event separately; slight differences in events. Discuss afterwards. Free recall of event.
71% witnesses included details they did not see.

More powerful than written misinformation

Question 17

Danger of interference

Answer 17

Natural logical inferences may act as self-generated misinformation.

• false memory for seeing schema-congruent slides that were not shown
• seeing effect slides inspired false memory for seeing ‘cause’ slides that were not shown

Casual inference errors affect memory within seconds.

Question 18

Exposure time and identification accuracy

Answer 18

Young and old witnesses view crime
Perpretrator visible for either
12sec or 45sec; asked to pick perpretrator out of target present or target absent line up 30 mins later.

Target-present
Young - 42% wrong
Old - 45% wrong

Target Absent
Young - 90% wrong
Old - 80% wrong

Question 19

False identification and weapon presence

What is weapon focus?

Answer 19

Presence of weapon = weaker identification, because weapon draws attention away from face and or contextually unusual objects draw attention.

Question 20

False identification and weapon presence

Evidence

Answer 20

Face fixation time is substantially lower with weapon presence

Threatening situations reduce identification rates anyway, but presence of a weapon does so even further; effects are modest.

Question 21

Ethnicity and Identification

Own-race bias or the other-race effect

Answer 21

Probability of identification error is about 1.5 times higher if target is different race compared to own-race.

An innocent black suspect is 55% more likely to be identified if witness is white as opposed to black.

Question 22

Ethnicity and Identification

Own-race bias or the other-race effect

Why?

Answer 22

Popular explanation is the contact hypothesis.

The more self-reported contact with opposite race, the higher the accuracy of identification.

Alternative explanation concerns social identity.

Faces perceived as in-group may be processed differently to those perceived to be out-group faces are processed as individuals.

Question 23

Facial Composites

Why use composites?

Answer 23

Verbal descriptions are impoverished and give a very low level of diagnosticity.

Composites allow a witness to illustrate their memory of the suspect’s face.

Question 24

Facial Composites

Are composite systems useful?

Answer 24

Composite systems vary widely; some evidence that some systems in use are inadequate

e.g. when students created composites of classmates, others in the same class could not identify the target. Zero identification rate.

Question 25

Facial Composites

Why?

Answer 25

People do not remember faces as a collection of features but remember faces holistically

Question 26

The problem with featural composites?

Are featural composites damaging?

Answer 26

Using featural composite may actually weaken later identification.

3 min exposure to face;
Composite group - make composite
Yoked-control - see composite made by previous ppt
Control group - neither make nor see composite 6-photo simultaneous line-up 2-days later.

Composite 
Target 10%
Filler 30%
No-choice 58%
Force-choice 30%

Question 27

Facial Composites

Why?

Answer 27

Does the process of using feature-based facial composite systems involve destroying the holistic links between features of the face?

Question 28

The identification line-up

Structure and function

Answer 28

Most line-ups are single-suspect, but can be multiple suspect.

Can be target-present (perpetrator is there) or target absent,

Given a line-up where the suspect is there, does the witness recognise the suspect?

Question 29

The identification line-up

What makes a fair line-up?

Answer 29

Functional size should match the actual size of line-up

Functional size is determined by number of members that are reasonable matches to perpretrator description.

Question 30

Using Line-Ups

How do witnesses identify perpetrators in a line-up?

Answer 30

What cognitive processes do witnesses use to complete the task?

What recommendations can psychology make to improve the accuracy and / or the utility of identifications

Question 31

Using Line-Ups

Relative judgement vs absolute judgement

Answer 31

Absolute: choose the candidate who is the perpetrator

Relative: choose candidate most like the perpetrator relative to other line-up members.

In simultaneous line-ups, witnesses most likely use a relative judgement.

Relative judgements lead to false identification in target-absent line-ups.

Question 32

The problem with relative judgement?

Relative judgement gone wrong

Answer 32

#3 is the perpetrator 
54% choose correctly when present; when perp is removed, other witnesses make a choice. 

If absolute judgement being used, what should happen?

For absolute judgements, expect 54% to make no choice.

Some do (no choice increases by 11%) but most go for the next closest match (poor #2).

Question 33

Two solutions to relative judgement

Pre-line-up instructions

Answer 33

the perpetrator may or may not be present.

Simple instructions reduce the proportion of false identifications in target absent line-ups

Question 34

Two solutions to relative judgement

Sequential line-up procedure

Answer 34

Many different variations sequential line-up e.g. multiple laps

Procedure is effective in reducing fale identifications in target-absent line-ups

Question 35

Sequential line up

Benefits

Answer 35

Reduction in false identifications in target-absent line-ups

Stage mock crime; sequential or simultaneous 6-person line-up; target present or target absent.

ID suspect 50% sequential
ID innocent 17% sequential
Not there 65% sequential

Correct identifications in target present around the same level.

Incorrect identifications in target-absent substantially reduced.

Question 36

Modern line-up in UK

Answer 36

Live line-ups are rare.

VIPER systems now common.
Foils chosen from over 20,000 person database.

Suspect and 8 foils shown sequentially 10 sec video of each person.

Line-up on DVD can be admisitered anywhere.

Witness asked to watch whole presentation before making any decisions.
Witness sometimes videoed making decision.