Reconstruction of Automobile Destruction: ! And Example of the Interaction Between Language and Memory!

!

! !

! !

By Elizabeth Loftus and John Palmer! Two experiments were conducted in which subjects were questioned about a film of a car accident they had watched. Subjects gave higher estimates of speed when asked “About how fast were the cars going when they hit into each other?” compared to subjects who were asked the same question but had the verb hit replaced with bumped or contacted. When they were tested a week later, subjects who received the verb smashed were more likely to say that they had seen broken glass, even though there was no broken glass in the film. These results support the idea that the questions asked after an event can cause a reconstruction in one’s memory of that event. !

How well can we remember the details of a complex event? It is well documented that most people are very inaccurate in reporting numerical details like time, speed and distance. Research indicates that people tend to overestimate the duration of complex events. In a test administered to members of the Air Force, they watched a car moving at 12 mph. Their estimates ranged from 10 to 50 mph. Given the inaccuracies in estimates of speed, it seems likely that there are powerful variables which influence these estimates. This research investigates how the phrasing of questions affect how an event is remembered. In this study, subjects were shown films of traffic accidents and then were asked questions about the speed of the vehicles.!

! ! Experiment I! !

Method! The sample consisted of 45 students who were divided into 5 groups. They were shown films depicting a traffic accident. They were then given a questionnaire which asked one groups “About how fast were the cars going when they hit into each other?” The other groups were questioned with the verbs smashed, collided, bumped, and contacted replacing hit. !

!

Results! Table 1 shows the mean speed estimates for different verbs. A statistical test showed that the words used had a significant effect on speed estimation. !

! ! ! ! ! ! ! ! ! ! !

Table 1. Speed Estimates fot the Verbs used in Experiment 1. Verb

Mean Speed Estimate (mph)

Smashed

40.8

Collided

39.3

Bumped

38.1

Hit

34.0

Contacted

31.8

Discussion! These results show that the phrasing of a question can significantly affect a witness’ answer. The actual speed of the vehicles had little effect in subject reporting, while the effect of the phrasing was significant. Two interpretations of this finding are possible. First, the difference in speed estimates may be due to response-bias factors. For example, a subject is uncertain whether to say 30 mph or 40 mph and the verb smashed biases his response towards the higher estimate. A second interpretation is that the question form causes a change in the subject’s memory of the accident. The verb smashed may change a subject’s memory such that he ‘sees’ the accident as being more severe than it actually was. If this was the case, we would expect subjects to recall other details that did not actually occur, but which would be associated with an accident occurring at higher speeds. The second experiment was designed to provide additional insights into the origin of the various speed estimates. !

! ! Experiment II! !

Method! The sample consisted of 150 students who were divided into 3 groups. A film depicting a car accident was shown, followed by a questionnaire about details of the accident. 50 subjects were asked “About how fast were the cars going when they smashed into each other?” Another fifty subjects were asked the same question but with the word hit instead of smashed. The other 50 were not asked about car speed. One week later, the subjects were asked the question; “Did you see any broken glass?”. There was no broken glass in the accident, but since broken glass is associated with high speed accidents, the subjects in the smashed condition were expected to say ‘yes’ more often. !

!

Results! The mean speed estimate for subjects in the smashed condition was 10.46 mph, with hit the estimate was 8 mph. These means are significantly different. Table 2 presents the distribution of ‘yes’ and ‘no’ responses for smashed, hit and control subjects. !

!

Table 2. Distribution of ‘yes’ and ‘no’ responses to the question, “Did you see any broken glass?” Response

Smashed

Hit

Control

Yes

16

7

6

No

34

43

44

! !

A statistical test showed that the verb used has a significant effect on the probability of saying “yes”. Subjects in the smashed condition were more likely to say ‘yes’ than subjects in the hit and control condition. Smashed leads to both more ‘yes’ responses and higher speed estimates. It was found that the the probability of saying ‘yes’ was not just an effect of the speed estimation made previously, the effect was still present even when speed estimation was controlled for. !

!

Discussion! We would like to propose that two kinds of information go into one’s memory for complex events. The first is information acquired during perception of the original event, the second is external information supplied after the event. Over time, information from these two

sources may be integrated in such a way that we are unable to tell from which source some specific detail is recalled. All we have is one ‘memory’. We propose that the subject first forms some representation of the accident that they have witnessed and then when the experimenter asks “About how fast were the cars going when they smashed into each other?” a new piece of external information is added, namely, that the cars did indeed ‘smash into’ each other. When these two pieces of information are integrated, the subject has a memory of an accident that was more severe than it actually was. Since broken glass is associated with severe accidents, the subject is more likely to think that broken glass was present. !

!