Loftus & Palmer - Eyewitness Testimony

Question 1

The leading critical question…

Answer 1

“How fast were the cars going when they SMASHED HIT BUMPED COLLIDED CONTACTED into each other?”

Question 2

Why was there a difference in verb in the leading question?

Answer 2

SMASHED suggests the cars were travelling quite fast, where as CONTACTED suggests a slower speed.

Question 3

Hypothesis

Answer 3

Strength of the verb will have a significant effect on participant reports of the speed of the crash

Question 4

Research question

Answer 4

Do leading questions distort an eyewitness memory of an event?

Question 5

How were order effects controlled?

Answer 5

Random sequence of presentation of films to each group

Question 6

What effect could demand characteristics have?

Answer 6

Student participants may work out the aim of the research

Question 7

IV experiment 1

Answer 7

The verb used

SMASHED HIT BUMPED COLLIDED CONTACTED

Question 8

DV experiment 1

Answer 8

The speed estimate

Question 9

IV experiment 2

Answer 9

Verb asked in the critical question

Question 10

DV experiment 2

Answer 10

Whether the participants recall seeing broken glass

Question 11

What were the two types of experiments?

Answer 11

1) memories at the time of an event occurring

2) memory after an event occurring

Question 12

Research method and design Experiment 1

Answer 12

Laboratory experiment

Between groups design
In the design of experiments, a between-group design is an experiment that has two or more groups of subjects each being tested by a different testing factor simultaneously.

Question 13

Sample Experiment 1

Answer 13

45 student participants

Question 14

Apparatus Experiment 1

Answer 14

7 film clips (5-30 seconds long)

Questionnaire - asked for the account of accident and answer specific questions

Question 15

Procedure Experiment 1

Answer 15

Students watched film clips of car accidents
After watching the films the students were asked to:
- write an account of what they had seen
- answer questions about what they had seen, this included the critical leading question

Question 16

How was the sample spilt in Experiment 1?

Answer 16

45 spilt into 5 groups of 9 people.
Due to 5 different verb conditions.
Changed order of films each time to stop order effect

Question 17

Findings experiment 1

Answer 17

Leading question did effect participants perception of speed

CONTACTED 31.8 mph LOWEST
HIT 34
BUMPED 38.1
SMASHED 40.8 mph HIGHEST
COLLIDED 39.3

These were average speed the participants estimated over 4 video clips. Only used 4 as more rigorous

Question 18

Conclusion Experiment 1

Answer 18

Phrasing effects speed estimate
Participants may mot be sure if is 30mph or 40mph, verb SMASHED biases them to respond with the higher estimate.
Alternatively the question causes a change in their memory, they may ‘remember’ detailed that did not occur - Expeirment 2 tests this

Question 19

Research deign and method of experiment 2

Answer 19

Lab experiment

Between participant design - different question for each group

Question 20

Sample experiment 2

Answer 20

150 students

Question 21

Apparatus experiment 2

Answer 21

1 film, 1 minute long, accident lasts 4 seconds
Questionnaire asking participants to describe accident in their own words and answer questions.
The time gap between when you witness and when you are asked, therefore makes it more realistic and life like.

Question 22

Procedure experiment 2

Answer 22

Film of multiple car accident presented - accident 4 seconds long
Questionnaire completed which included the critical speed question.
3 different critical speed questions. “How fast the cars were going when SMASHED, HIT” or… NO Q.
50 in each group.
One week later (to make more realistic) participants were asked 10 questions.
Randomly placed critical question asked “did you see any broken glass.” YES, NO.
- there was no broken glass but it does fit with high speed accidents, so expect participants to say yes more often with SMASHED verb than HIT verb
“The broken glass” implying there was some

Question 23

Findings experiment 2

Answer 23

Mean speed estimate more for 
SMASHED 10.46mph... HIT 8mph
Probability yes to broken glass 
SMASHED 0.32
HIT 0.14
   Participants did give more yes responses and higher speed estimates for smashed
    SMASHED 16/50
    HIT 7/50
    CONTROL 6/50

Question 24

Brief summary procedure experiment 2

Answer 24

Film 1 minute 4 sec accident
Hit or smashed for post info question
Did you see broken glass? To see if post info effects

Question 25

Conclusion experiment 2

Answer 25

Meaning of verb used in the leading question had become integrated with the memory of the event, thus changing the memory to be constructed.
Can conclude what happens after we have witness an event can alter our memory of the event.

Question 26

Conclusions whole study

Answer 26

Post event info can change memory of the event eg. Leading question.
Severity/intense nature of the word used can effect the estimate.
Phrasing a question influences recall.
2 types of info go into memory:
- one from perception of original event
- one from external information supplied after leading question
These merge over time, so we experience effects on memory.
Verb SMASHED labels the event causing a shift in memory.

Question 27

Strengths of method

Answer 27

Manipulated words in questions and measured effect on recall in lab.
1: controlled conditions= keeping many controlled as possible, therefore could conclude that it was the words used on the questions which caused difference in recall.
2: films participants saw= everyone saw same films in same room and questioned about in same way (with exception of change in verb IV)
3: asked same questions immediately after films shown, therefore no variance on the way questions were asked and any feedback from experiment was possible.
MAJOR STRENGTH = CONTROL
the more variables you have control over the more you can draw conclusions from IV & DV.
CONTROL OVER… critical question, films and questionnaire.

Question 28

Weakness in method

Answer 28

• High levels of control = artificial situation, therefore difficult to imply results to everyday life, therefore LOW ecological validity
• Participants know they are taking part in lab experiment, could alter behaviour…
  Look for clues on how to behave (demand characteristics)
  Want to help experiments by giving desirable results
  Therefore major effects on results of the study

Question 29

Representativeness of sample

Answer 29

Students not representative of whole population
Therefore difficult to generalise to whole population and people in general

Differences may be:
Students are young, and possible memory is better when young
Students used to taking in lots of info no being asked questions about it. People who haven’t studied for many years may have more difficulty with this.
Students less experienced drivers and therefore less able and confident to estimate speed. Therefore more influenced by verbs in the leading question.
Students may be more susceptible to demand characteristics, especially if students of researchers who are conducting the study.

Question 30

What type of data collected?

Answer 30

Quantitative: speed estimated and no people who has seen broken glass.
Useful for comparison and allows statistical analysis, but fairly superficial and does not tell us anything about why people gave answers they did.

Could improve by asking how confident they were in answers they had given, would allow more detailed analysis.

Question 31

Usefulness of research

Answer 31

Conclusion that leading question can affect memory has important implications for interviewing witness, both police immediately or soon after an event. Also by lawyers in court sometime later.
Interviews should avoid leading questions and show be careful to word questions in ways that do not suggest an answer to the person they are interviewing (do not want bias).
121 answered no correctly in experiment 2, including 2/3 in the smashed verb condition. Therefore proves not so easy to change memory with critical question.
Also if misleading info is blatantly incorrect people are less likely to take in and overwrite previous info.

Question 32

Ethics

Answer 32

Clips of crashes from safety films, therefore mo gruesome images - should not have upset subjects or caused harm.

Ethically worthy as contributes to debate about witnesses to events and how they should be questioned in order to get a true picture.

Not ethical to conduct experiment in real life setting, would be dangerous and risky. Could cause injuries and psychological harm.

Question 33

Ecological validity

Answer 33

LOW
Conducted in lab
Different watching a safety video compared to an actual crash.
In real life there would be other distractions and high emotional involvement.
Accidents happen spontaneously in real world and memories may be recalled differently under distressed situation compared to watchi a video clip.

Question 34

Validity

Answer 34

Challenged to external validity = low generalisability (student sample).
Reliability = findings support reliability that language (leading questions) can distort memory.
Concurrently valid between studies as findings of experiment 2 agree.

Question 35

Aim

Answer 35

To see the effect of leading questions on memory specifically to find out if changing a verb used in a question about speed could effect the speed given.