Loftus and Palmer | Tony Ryan's Web Site

Elizabeth Loftus

Elizabeth Loftus (born 1944) is a very famous psychologist and was voted as one of the most famous cognitive psychologists in the world. Her main are of expertise was human memory and how it can be manipulated. In the study that I am about to describe she investigates how the phrasing of a question can influence the memory of the event the question was about.

So, let us look at the study: If you want your own copy then the reference is as follows:

"Reconstruction of Automobile Destruction: An example of the Interaction Between Language and Memory. Elizabeth F. Loftus and John C. Palmer. Journal of Verbal Learning and Verbal Behaviour. 13, 585-589 (1974)

It is quite an easy read, and well worth the effort.

The first question to answer is why the study was carried out in the first place?In certain situations it is important that we need to remember the details of a complex event, such as a car crash. Before this study, it was well known that people overestimate the length of time that a complex event took.

It was also well known that people have a great deal of difficulty estimating the speed of a car. U.S. Air Force personnel were asked to estimate the speed of a car. Estimates varied between 10 mph and 50 mph. The car was going at 12 mph.

Given that estimating things is so difficult, are there things that might influence these estimate? Loftus and Palmer were investigating whether the words used in the question influenced the answer about the vehicle's speed.

Another concept that is relevant to this area is the concept of a leading question. This is a question that implies an answer by the words used in the question.

The Study.

The method used:

The method used in both parts of this study was "laboratory experiment". It was an experiment as it had an Independen Variable that changed, and it was carried out in a laboratory.

The Procedure:

The "Procedure" differs from the "Method" as the procedure is what happened, and the method is how it was done.

So we have two experiments in this study, the method used in either was the same, but the procedure differed.

Experiment 1

Students participated in groups of various sizes. They were shown seven films depicting car crashes. Following each film the students were asked to write down what they had seen. Then they were asked specific questions about elements of the film. The critical question was about the speed of the collision. Nine students were asked "About how fast were the cars going when they hit each other? Equal numbers of students were asked slightly different question with the verbs smashed, collided, bumped and contacted in place of hit. Films were shown in a different order to the different groups of students.

Experiment 2

Students participated in groups of various sizes. A film depicting a multiple car accident was shown, followed by a questionnaire. The film lasted less than 1 minute and the accident in the film lasted for 4 seconds. After the film the students were asked to describe what they had seen, in a questionnaire. Fifty students were asked "About how fast were the cars going when they smashed into each other" Fifty student were asked the same question, but smashed was changed to hit.

A week later the students returned and, without seeing the film again were given another questionnaire about the accident they had seen a week previously. There were 10 questions and the critical question was "Did you see any broken glass?" The students responded by checking a box for "Yes" or "No". The order of the question were randomised, and there was no glass visible in the film.

The Samples used.

In Experiment 1 there were 45 students. In Experiment 2 there were 150 students. All students were probably students at the University of Washington, where Elizabeth Loftus was the Professor of Psychology.

The Findings (Results)

In Experiment 1 the results were as follows:

Speed Estimates for the Verb used in Expt, 1

Verb Mean speed estimate

Smashed 40.5

Collided 39.3

Bumped 38.1

Contacted 31.8

The results showed that compared to random, these results were "significant" at 0.05 level.

Some of the film showed "Staged crashes" where the speed of the vehicles was known. The students estimated the speed of the one car travelling at 20 mph as being 37.7 mph. The estimate of the one car travelling at 30 mph was 36.2. The two cars travelling at 40 mph were

estimated to be going at 39.7mph. This was not a significant finding but showed that estimates of speed is not something people are good at.

Conclusions from Experiment 1

From the Findings above it is possible to conclude that speed estimates of the cars involved in a crash can be influenced by the verb used in the question. There is a clear difference in the estimate from those students who heard "smashed" in the question to those that heard any other verb.

There are two possible explanations:

1. The difference may be due to "response bias". This is where the student is unsure whether to say 30mph or 40 mph and the verb they hear in the question biases them on way or the other, e.g. a student is more likely to guess 40 mph when hearing "smashed" but 30 mph when hearing "contacted".

2. The second explanation may be that the different verbs used in the question may interfere with the way that the crash is remembered, e.g. a student who hears "smashed" may remember the crash being more severe than if he heard "contacted".

In Experiment 2 the results were as follows:

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

Analysis of these results show that there is a significant difference between the answers to the "smashed" question and the answers to the "hit" question when they answered "Yes" to "did you see any broken glass?"

It was also found that those that heard "smashed" in the question estimated the speed of the cars to be higher that those that heard the "hit" question. That lead onto thinking about the effect of hearing the word "smashed" when being asked about car crashes.

Conclusions from Experiment 2

From the findings above, it is possible to say that this study has proved that the responses given to a question may be different, depending on what verb is used in the question.

A framework to discuss the changes that might cause these differences is that it might be that our memories of events are made at the time, but further information about the event may corrupt the original memory of the event. It is thought that we only have one memory of one event. If further information is added to that memory we won't be able to tell which bits of our memory came from where. This is called the Reconstructed Memory Hypothesis. Evidence of this comes from the fact that in experiment 2 the critical question came a week after they had seen the film of the car crash.