Support and refutation of a threat-inducing communication in change of attitude toward cigarette smoking.

Abstract

A central problem of research on persuasive communications has been whether a prior communication has any effect in reducing the effectiveness of a later communicauon (Hovland, 1957). The present study reports the effects of support and refutation of a prior threat-inducing communication on change in attitudes toward cigarette smoking. Of the 86 freshmen and sophomore university students who took the modified Buer Scale for Measu~ing Attitudes toward any Practice, Form A (Bues, 1934), 28 Ss were classified as Heavy Smokers (HS, 20 cigarettes/day) and 25 Ss as Light Smokers (LS, 10 cigarettes or fewer/day). Ss in each group were randomly assigned to Supportive (S) or Refutative (R) type communication treatment, making four experimental groups, HS-S, HS-R, LS-S, LS-R. Seven days after taking Form A of the Bues Scale, all Ss saw a 16-mm. film prepared by the American Cancer Society and entitled Is smoking worth it? The threat-inducing communication of this film was that cigarette smoking is harmful. Following this a supportive (Group S) or refutarive (Group R) speech was delivered. There were 10 Ss in each of the four experimental groups. After the speech all Ss took Form B of the modified Bues Scale. Two control groups (10 HS and 10 LS) merely took Form A and Form B of the Bues scales, with the same intertest interval of seven days. The Pearsonian correlation for the test-reliability of the two parallel forms of the attitude scale was .85. A Lindquist, Type 111, mixed analysis of variance of raw attitude scores indicated a significant decrease in attitude scores, i.e., a more negauve attitude toward smoking, for all six groups (P = 26.72, df = 1/54, p < .001). No interactions were significant, i.e., there was no differential attitude change toward smoking for HS or LS receiving supportive or refutative communications. An additional analysis showed that a change in attitude scores for the 4 speaker groups was greater than for the 2 control groups (t = 4.42, df = 54, < .01). There was only a trend for the HS group to express a more negative attitude toward smoking under S-treatment and for the LS group to express a more negative attitude under the R-treatment. It was concluded char the presence of the speaker and/or the film had a significant effect on the change in attitudes toward smoking; however, the type of communication following threat-induction did not.