loftus and palmer - classic study Flashcards
experiment 1 aim
to determine whether the intensity of a verb used within a question asking about the speed of the cars in an accident altered the speed estimate of those who watched the accident on film
what experimental design was used -1
independent measures
the 5 different verbs used - 1
smashed, collided, bumped, contacted and hit
sample - 1
45 students
lacks generalisability as they have limited driving experience
details of the film footage - 1
7 5-30 second long films depicting traffic accidents
drivers education films borrowed from the Evergreen Safety council and Seattle Police Department
different ordering of the films presented to each group - counterbalancing
4 staged accidents - can control the speeds the cars were going at
procedure - 1
7 films ranging 5-30 seconds of road traffic accidents shown to ppts in random order
received questionnaire after each film which asked them to give an account of what they had just seen and asked a series of questions (including critical question)
critical question - 1
‘about how fast were the cars going when they hit each other’ - with either the verb smashed, hit, contacted, collided or bumped
what was the mean speed estimate for the verb smashed? - 1
40.5 mph
what was the mean speed estimate for the verb contacted? - 1`
31.8 mph
what was the issue with only having 9 ppts in each group? - 1
any anomalies would drastically skew data
what do the results suggest - 1
the form of a question can affect a witness’s answer to that question
the more powerful the verb was, the higher the speed estimate
conclusions - 1
distortion - verbal label attached to the event could have cognitively changed their memory
response bias (demand characteristic) - not sure of the exact speed so they adjust their estimate to fit with the expectations
aim of experiment 2
to determine whether the speed estimates were due to memory distortion or response bias
sample - 2
150 participants
divided into 3 equal groups
procedure - 2
a film lasting less then 1 minute with a 4 second car crash was shown to 150 participants
ppts then received a questionnaire where they first had to describe the accident in their own words then answer a series of questions - critical question was, ‘about how fast were the cars going when they hit each other?’ with the 3 conditions being smashed, hit, and not interrogated about speed.
one week later the subjects were asked 10 questions about the film with the critical question being - ‘did you see any broken glass?’ - no broken glass in the video
results - 2
ppts in the smashed condition gave significantly higher speed estimate than those in the hit condititon - 10mph and 8mph - supports first exp.
probability of saying yes is 0.32 when smashed is used and 0.14 with hit
smashed leads to more yes responses and higher speed estimates
experimenter bias - only uses yes data when analysing results - ignores no data
conclusions - 2
memory of an event was changed via leading questions rather than just a response bias
two types of information go into one’s memory - information gained during the perception of the original event and information supplied afterwards
over time, the information from these two sources become integrated, so we are unable to tell from which source the detail is recalled, as we only have one memory
the label smash causes a shift in memory representation of the accident in the direction of being more similar to a representation suggested by the verbal label
supporting research
Daniel (1972) - showed that recognition memory, as well as reproductive memory, was similarly affected by verbal labels, and concluded that the verbal labels cause a shift in the memory strength to forms which are better representatives of the label
Filmore (1971) - noted that hit and smashed may involve specification of differential rates of movement
opposing research
Yuille and Cutshall - (1986)
- misleading information did not alter the memory of people who had witnessed a real armed robbery
- meaning that misleading information may have greater influence in the lab
G
P - low
E - 45 students in exp 1 were all students
E - students all have limited driving experience, so the sample isn’t representative of people with different levels of driving experience
P - good
E - sample is ethnocentric, but memory is universal
E - the results should be the same across all countries and cultures as memory is the same for everyone
R
p - high as good standardised procedure
E - 7 5-30 second long films (4 being staged) were shown to each group
E - the study can be replicated and tested for consistency as the procedure was closely controlled, meaning that variables can be controlled for if the procedure is repeated
P - high as quantitative data is collected
E - mean speed estimates and ‘yes’ or ‘no’ answer to the critical question ‘did you see broken glass’
E - quantitative data is objective and it prevents researcher bias as subjectivity doesn’t have to be introduced to analyse these results
A
P - leading questions can impact the recall of an event
E - smashed resulted in a speed estimate of 40mph whereas hit was 34mph in exp. 2
E - police should be advised to avoid leading questions as any verbal label in the question can lead to distortion of the event, which can cause an eye witness testimony to become less accurate
P - could be used by lawyers in the court room
E - could discredit eyewitness testimonies as leading questions and response bias may influence them
E - as this research shows that their memory can be altered after the event,
V
P - low ecological validity
E - lab experiment, watched 7 5-30 second long films of car crashes
E - a witness would normally only see one event at a time, whereas these ppts have witnessed 7, which means they won’t have as much emotional involvement and attention as an actual witness would.
P - high as independent measures design used
E - 9 ppts asked for each verb, hit, collided, smashed, bumped and contacted in exp. 1
E - prevents demand characteristics as ppts are less likely to guess the aim of the research
P - high internal validity due to counterbalancing
E - the ppts were shown the same clips but in different orders
E - practice effects won’t impact the overall results, as a different video would have been shown last for each group
E
P - good as ppts protected from psychological harm
E - 4 of the videos were of staged car crashes
E - less distressing as the ppts know that no one was harmed in the making of the video or for the purpose of the experiment.
