Brodt, S. E., & Zimbardo, P.G., Modifying Shyness-Related Social Behavior Through Symptom Misattribution, JPSP, 1981,41,437-449.

This study attempted to improve the shyness of women subjects by having them attribute their shyness to an unrelated cause.

A pattern of anxiety-evoking stimuli that subsists over time and becomes part of a person's "being" can be called a disposition. Often the assumption is that the disposition is self-motivated rather than a response to a stimulus. Often the person feels helpless to change this "disposition". Their anxious responses to certain situations can set up a stereotype that is reinforced through their behavior.

"Both the perception of arousal and it's explanation are crucial to the predicted flow of the unfolding process (i.e., the process that produces the unwanted, stereotyped behavior). If you can explain away the physiological change, the unfolding process may not happen.

One way to "head off" the unwanted behavior is through arousal misattribution. This is accomplished by attributing the arousal to another neutral stimulus in the environment (not the real cause). The subject is led to believe that their unwanted actions are an unconscious result of this "other" stimulus. When the person is put in the usually arousing situation, any reactions are explained away as the placebo factor. Thus the undesirable factors are not provoked and thus avoided.

The researchers did their work on self-proclaimed, extremely shy women, especially shy when with the opposite sex. They were led to believe that their shyness was actually caused by high-frequency noise. They were told that shyness was a common "side effect" to exposure to high frequency noise. The researchers predicted that the next time they experienced this social situation (which should make them feel shy), they would relable their arousal as caused by a non-psychological source (noise) of which they bore no responsibility. By breaking the normal causal relationship, shyness would be reduced.

They compared three groups -- one shy groups who received the misattribution, one shy group that didn't, and one non-shy group that recieved the misattribution.

They selected Stanford students and classified by a shyness test. The shy women were joined with a man (confederate) and they were told the experiment was on the effects of noise pollution on physiological responses. They were told the noise may produce increase heart rate, dry moutn, and slight tremors.

They hooked up equipment and did some fake testing, and then left them alone for a "waiting" period. The male confederate then asked some questions from a script and allowed the subject to respond freely. The confederate was blind to the professed shyness levels.

They scored the conversations between the two, some self-reports after the study on level of anxiety, and physiological changes by machine.

Results
Most subjects did notice the heart rate increase and poinding warned about in the misattribution (shy and not-shy). The shy control group (who didn't receive the misatttribution) didn't notice the heart rate but experienced tremors and dry mouth (consistent with being shy).

Most of the subjects in the two misattribution groups had negative self-reported feelings about the noise. None of the shy control group did.

The misattribution did significantly improve the talkability of the shy misattributed subjects vs the shy control. In fact, the misattributed shy group was similar to the non-shy group.

The confederate (blind to their shyness condition) more or less correctly identified the not-shy people and the shy control group. But he misjudged the shy misattributed group.

In a post questionnaire, the shy misattributed group liked the "togetherness" of the experiment as opposed to the shy control group.

Conclusion
The results suggest that social behavior can be altered by misattribution. Shy people, given an alternate explanation for their shyness, became less shy.