Across 2 studies, the authors explored reactance effects to overexaggerated claims and controlling language in exercise messaging. In Study 1, participants received either a message exaggerating the benefits of an upcoming exercise session or no message. They subsequently undertook a mundane exercise session led by an instructor, which was either need supportive or “realistically controlling.” Relative to no-message participants, those who had read the message reported less positive evaluations of the session. These results were observed despite participants in the message condition holding more positive presession expectations, and the effect was apparent even for those who received need-supportive instruction. In Study 2, participants read an advertisement that was written in either autonomy-supportive language or controlling language. Despite reporting comparable expectations, participants who received a controlling-language message reported significantly greater anger and freedom threat—factors commonly linked to contrast effects. These studies highlight the operation of message-driven contrast effects in exercise.
Dimmock, Simich, Budden, and Jackson are with the School of Human Sciences, University of Western Australia, Perth, WA, Australia. Podlog is with the University of Utah, Salt Lake City, UT. Beauchamp is with the University of British Columbia, Vancouver, BC, Canada.