Elizabeth Loftus is a memory expert affiliated with the University of Washington and the first to identify and research the effect of interference on memory. Loftus discovered that memory can be manipulated so that future perceptions and remembrances differ from original perceptions and memories. These interference effects cause subjects to lose their memory representations of what really happened.
Loftus found that subjects' memories could be manipulated so that they would recall seeing things they had not witnessed. Loftus found a way to plant wrong ideas at three different phases of memory: perception, storage and retrieval. The repercussions of Loftus' work are that a child may, for example, "remember" having been the victim of abuse after interference by psychologists or others, though the abuse may not have occurred.
The work by Elizabeth Loftus began with studying how phrasing a question a certain way might change the subject's response. In one study, subjects were asked to watch a filmed automobile accident. The subjects were then asked a series of questions. Those subjects who were asked, "Did you see the broken headlight?" were more likely to answer in the affirmative than if they were asked, "Did you see a broken headlight?" There was no broken headlight depicted in the film. But by using the word the, an implication was made that there was, in fact, a broken headlight. Therefore, subjects responded as if they remembered the nonexistent broken headlight.
A similar experiment involved asking some of the subjects in a research trial: "Do you have frequent headaches and if so, how often?" Subjects reported that they averaged 2.2 headaches each week. Then Loftus asked the same question of other subjects, substituting "occasional" for "frequent." These subjects reported that they experienced just 0.71 headaches each week. These subjects were assigned to their groups at random. It is unlikely that the first group experienced so many more headaches per week than the subjects in the second group. Therefore, it seems that using the word frequent as opposed to the word occasional generated a response signifying more frequent headaches. This is suggestive that the use of the word frequent presupposed that the number of headaches would be high.
Loftus' novel idea was that a person's responses are influenced by the manner in which questions are worded. The observation is an important one. However, this observation failed to tell us the mechanism by which the "interference effect" is transacted.
For example, it may not be the wording of the question influencing responses, but rather the perceptions of the subjects that the questioner wanted a certain response. When the researcher asked, "Did you see the broken headlight?" the subjects might have assumed that the researcher wanted them to have seen the broken headlight. The subjects wish to cooperate with the desires of the questioner. There is also the possibility that the manner of questioning causes subjects to begin to doubt the truth of their real memory. The subjects fear having missed something crucial and so change a response to allay such fears: Had they been paying attention, they might have seen that broken headlight. Through fear or suggestion or some other reason, the representation of the event is changed in some fundamental way.
In order to run down the various possibilities of why subjects changed their perceptions and memories, it was necessary to see how subjects responded to different questions. In 1974, Loftus and colleague John Palmer showed subjects a film of a car accident. Some subjects were asked, "About how fast were the automobiles going when they smashed into each other?" The subjects responded with an average estimate of 10.5 miles an hour.
Another group of subjects were asked the same question with the word hit used instead of smashed. The more neutral verb elicited an average estimated speed of just 8 miles an hour. Thus far, this only repeats the earlier study in which a particular response is generated through wording. However, something novel then occurred.
The subjects were recalled one week later and asked further questions relating to the film. This time, the main question was: "Did you see any broken glass?" though there had been no broken glass in the film. The subjects who had been asked the earlier question employing the verb smash were significantly more likely to remember having seen broken glass in the film compared to those who had been questioned using the more neutral verb hit one week earlier.
The conclusion of this research is that memory can be altered through information received after the fact. A transformed memory can depict a more serious incident than the one that actually occurred. When the language of the questioner is not strictly neutral, the memories of the respondent can be changed forever, irrespective of actual events.