Can the public expectation of emotion cause that emotion?

An experiment investigated reactions to expectancies as a function of the reasons from which the expectancies were derived. Each subject was told that he was expected to feel inhibited performing a singing task. The expectancy was presented as based either on the subject's description of his personality, on the past performance of others with interests similar to those of the subject, or on the past performance of others with the same birth order position as the subject. Subjects then sang a piece without accompaniment for a tape recorder, ostensibly providing data about the effects of inhibition on the physical properties of the human voice. Subjects expected to be paid proportionally to the duration of their singing. The expectancies based on self‐descriptions and on others with similar interests elicited faster singing, implying a willingness to sacrifice financial rewards in order to end an embarrassing situation, than the singing of no expectancy control group subjects, suggesting that these subjects actually did feel more inhibited than the control subjects. The expectancy based on birth order did not produce singing durations that differed significantly from the control group. The findings are interpreted as implying that persons will come to feel the way they are expected to feel only if the expectancy is perceived as deriving from some characteristic reflecting free choice and control.