Loftus and Palmer core study

study 1: effect of leading questions on ability to accurately remember an event

study 2: to see if the results of study one were due to response bias or leading questions distorting the memory of the original event

• opportunity sample of students
• from university of Washington
• study 1: 45 participants, 5 groups of 8
• study 2: 150 participants, 3 groups of 50

Study 1

• participants in each group watch 7 film clips of car crashes (originally from a driver safety film) in a different order
• complete 7 questionnaires (one on each film clip)
• includes the critical question 'About how fast were the cars going when they .......... each other?'
• verbs included contacted, bumped, hit, smahed, collided
• took 90 mins to complete the experiment

Study 2

• participants watched one film clip
• completed one questionnaire
• included the same critical question about speed estimate
• used the verbs hit, smashed and included a control group with no question
• 1 week later, participants were asked another critical question 'Did you see any broken glass?'
• there was no broken glass in the original clip

Study 1

• participants with the verb 'smashed' in the critical question, had the highest mean speed estimates- 40.8mph
• participants with the verb 'contacted' had the lowest mean speed estimate- 31.8mph
• overall, participants speed estimates of the vehicles were not accurate to the original video

Study 2

• participants with the verb 'smashed' had a higher mean speed estimate (10.46) than those with the verb 'hit' (8.00)
• those with the verb 'smashed' answered yes more frequently to the critical question about broken glass (16/50) compared to with the verb 'hit' (7/50) and the control group (6/50)

Study one

• P: leading questions have an effect on estimate of speed
• E: this can be seen in the study where those with the verb 'smashed' had the highest mean speed estimate of 40.8
• E: therefore, participants overestimated speed due to distorted memory and use of leading questions

Study two

• P: high speed estimates were due to leading questions distorting memory rather than response bias
• E: this can be seen in the study where those with the verb 'smashed' answered yes to the critical question about broken glass more than the group with 'hit' or the control (16/50 compared to 7/50 and 6/50)
• E: this therefore shows how memory is reconstructive and possible schemas were activated

Change 1

• Description

change the sample to include other people besides students. this could be done by placing advertisements in a number of local newspapers explaining about the study of memory. collect a sample of 100 participants from the advert- 50 males, 50 females, above the age of 25. screen participants to ensure they have over a year of driving experience

• Advantage

more representative of the wider target population- larger age range

• Disadvantage

time consuming and potentially expensive to carry out, difficult to recruit 100 more participants who are willing and different to the 195 original participants

• Implications -findings are more generalizable, results may include speed estimates that are more accurate to the original film clips

Change 2

• Description

change from a lab experiment to a field experiment. to ensure the study is ethical, conduct a staged car crash that people in the public witness. ask eye witnesses to complete a questionnaire, including the same questions as that of the original study and the critical question about speed estimate and broken glass

• Advantage- higher ecological validity
• Disadvantage

less control of variables i.e. number of people in the public that see the crash, not all those who see it may want/get asked to complete the questionnaire, lower reliability

• Implications- more confounding variables, however also more valid- there may be some emotional response associated with ability to accurately remember an event- believe the staged crash is real