Psychological Reactance and Persuasive Health Communication: A Test and Extension of the Intertwined Model

Abstract

This manuscript reports 2 experiments that were conducted to test and extend the work of J. P. Dillard and L. Shen (2005) examining the cognitive and affective processes involved in psychological reactance. In particular, the studies reported here (a) examined the best-fitting model of reactance processes and (b) tested 3 factors that may affect reactance including argument quality, severity of the consequences associated with the message topic, and magnitude of the request made in the message. The results showed that the intertwined cognitive-affective model was the best-fitting model of reactance processes. Magnitude of the request was the only variable that affected reactance. The implications of these findings for research on reactance and persuasive health campaigns are discussed.

When Artificial Social Agents Try to Persuade People: The Role of Social Agency on the Occurrence of Psychological Reactance

“…Different from the effect of social agency on the psychological reactance measures, which was only found on the negative cognitions measure, the effect of threat to autonomy of the language in the advice was found both on the measure for negative cognitions and the measure for feelings of anger. This finding is in correspondence with earlier research [8,17,19,21,23] that indicates that the use of threatening language (threatening people's autonomy) and controlling language (using explicit directives) will diminish persuasion and enhance psychological reactance. Moreover, this finding was especially pronounced on the reactance measure of feelings of anger, but also present on the reactance measure of negative cognitions.…”

“…Furthermore, we expected an interaction between threat level and social agency level, leading to the strongest effect of psychological reactance when high threat would co-occur with high social agency (hypothesis 3). Finally, in line with previous research [8,21,23], we expected a positive relationship between threat to autonomy and restoration intentions, and that this relationship would be mediated by psychological reactance (hypothesis 4).…”

“…In their study, they present two concepts that could be used as indicators of psychological reactance; anger and negative cognitions. Further research showed that these two factors could be explained best by an intertwined model-the two indicators contributed to psychological reactance in an intermingling way [8,21,23]. That is, anger and negative cognitions were best explained as an inseparable construct.…”

“…[6,8,9,17,19,[21][22][23]25]. This research indicates that when language becomes more controlling, people are more prone to respond in 1 The original theory of psychological reactance [3] speaks of perceived freedom instead of perceived autonomy.…”

“…That is, results indicated that participants who read a high-threat message had a more negative attitude towards the advocated behavior (e.g., responsible drinking) in the message, and had lower behavioral intentions in line with the advocated behavior (e.g., responsible drinking) than participants who read a low-threat message. In addition, it was found that this relationship was (almost) fully mediated by anger and negative cognitions [8,21,23]. …”