= 389) were evenly distributed across gender, education, and age. Structural equation model results indicated that the theories complement-rather than compete-with each other, as theoretical mechanisms between language and intentions were supported for both theories and some associations merged theoretical predictions. Participants who perceived messages to be freedom threatening experienced increased reactance and therefore had reduced positive attitudes about the behavior and reduced behavioral intentions. Participants who evaluated messages as having more desirable characteristics, specifically positive facework, negative facework, efficacy, and feasibility, experienced increased behavioral intentions, either through direct associations or by also having reduced reactance or increased positive attitudes about the behavior. Indirect effects occurred between the language manipulations and behavioral intentions through perceptions of positive facework, negative facework, and freedom threat, but surprisingly differences only emerged between mitigated and aggravated face-threatening language, with no distinctions between whether the language centered on positive or negative face. We discuss theoretical and practical implications, including merging theoretical considerations of face-threat, efficacy and feasibility evaluations, reactance, and attitudes to maximize predictions of behavioral intentions.
Guntzviller et al. (2026) studied this question.