Friday, June 6, 2008

Reconstruction of Automobile Destruction. Loftus, E. F. and Palmer, J. C. (1974)

A review of the experiment by Loftus, E. F. and Palmer, J. C. (1974)

To investigate the influence of information supplied after an event on a person’s memory of the event.

Two separate, though similar, experiments were conducted.
Experimental Design: Independent measure design
Independent Variable: Verb used
Dependant Variable:
For one of the experiments (Experiment 1), participants’ speed estimate.
For the other experiment (Experiment 2), belief of the participants that they saw glass

Experiment 1:


To investigate how the information provided after an event affects the person's memory of that event.

Materials used:
45 participants (students of University of Washington)
7 film-clips of traffic accidents ranging from 5 to 30 seconds (excerpts from safety films made for drivers’ education)
Students were made to write an account of the accident clip.
They were made to answer questions. The wordings of one of these questions, which was related to speed, were manipulated (independent variable)


The experimental group was divided into five.
Each group was subjected to different conditions; i.e. the independent variable was manipulated, so that the questions posed for each group had one of the words from ‘smashed’, ‘collided’, ‘bumped’, ‘hit’ and ‘contacted’ (words showing different intensities of impact) inserted into the question ‘About how fast were the cars going when they ______ into each other’
The ordering of the films was also manipulated, such that every person in 1 condition group was shown the clips in a different order.
The speed estimates of the participants were recorded.


The following chart displays the difference between the estimates provided by participants in the different conditions.
It can be seen that a variation 9 mph occurs between the estimates provided, from 1) smashed, the verb with most intensity to 5) contacted, the verb implying least intensity

Loftus and Palmer explained these findings by saying that
1. The verbal label provided by the experimenter for the accident distorted the memory of the participants
2. They confess that the results could be due to ‘response-bias’ meaning that because they did not estimate the speed with surety they could have adjusted the speed according to the condition.

Experiment 2:


The second experiment was conducted to investigate why information
supplied after an event influences a person’s memory of that event.

Materials used:
1. 150 participants (students)
2. A minute-long film which contained a 4-second sequence of an accident involving more than one car.
3. The participants were interrogated about the clip, with the manipulation of the IV occurring with the experimenter using the word 'hit' or 'smashed' in the question 'How fast were the cars going when they ______ each other?'
4. The participants were asked, a week later, several questions. The one that determined the DV was 'Did you see any broken glass?'

· The participants were divided into three groups; two experimental and one control.
· The IV was manipulated for the experimental groups when they were asked questions about the experiment. The control group was not interrogated.
· The participants were asked to return a week later. They were, then, asked some questions about the clip.
· The critical question about the broken glass was asked randomly and the responses of the participants were recorded.

The following chart shows how the responses of the participants varied with the independent variable:
The results show that the number of participants from the 'smash' group who believed that they had seen glass is significantly higher than the 'hit' or 'control' group.

Based on these findings, Loftus and Palmer developed an explanation which says that the information going into the memory is of two kinds, the information obtained from perception of an event and that obtained from information supplied after the event.

Over time, the information gets integrated such that one is not able to differentiate between the sources of the details. In the end, people only have one memory. This is called 'reconstructive hypothesis'.

In the second experiment, the memory of the video-clip and the word 'smash' which implied that the cars did smash into each other. These two memories become integrated and the participant is led to believe that the accident was more severe than it actually was.


The experiment was well-controlled as Loftus and Palmer had controlled the timings and place of the experiment. Moreover, varying the pattern of viewing the clips for participants and placing the questions in the second experiment randomly makes the experiment more valid. Because films were shows, it provided a filter for participant-bias.

However, the use of the findings to prove that speed estimates provided by witnesses cannot be considered as valid. This is because the environment in which this study takes place and the environments in real-life situations are different. This means that the study is not quite ecologically valid.

Moreover, there is a difference in the areas of interest and memory capabilities of students and other people.

Basically, the findings face criticism on their being generalized. For example, memory of eye-witnessed events could also be influenced by factors such as sleeplessness, alcohol, emotional attachment and the environment etc.

Another criticism faced by these findings is that the responses were either biased because of the questions asked, or that the queues provided by the questions led to a distortion in the retrieved memory (the original memory remained intact)

1 comment:

Anonymous said...

What is the research hypothesis?