Portions of this work have been previously presented at the Society for Personality and Social Psychology annual conferences (2017, 2018, 2019) and the associated pre-conferences on Justice and Morality and Law and Psychology. Additionally, separate analyses of these data are reported in a companion manuscript (Celestin & Kruschke, 2019). Although the two manuscripts are based on the same data, the analyses are (a) mathematically different and (b) focus on very different aspects of the models. The latent severity scales reported here remain essentially unchanged when using more elaborate models, so we present the most parsimonious model here in order to maximize the clarity of the latent scale derivation and the implications of those results.

Most police agencies in the United States rely on a rank-ordered categorization of forceful actions, along with decision heuristics for appropriate force options in response to civilian actions. We had lay people rate the moral and physical severity of officer reactions to civilian actions, and vice versa. The actions were representative of the entire range of force options in official police policy and realistic police-civilian conflicts. Results showed that ratings of moral severity of a reaction decreased as severity of the initial action increased, and that ratings of both police and civilian reactions were modulated by individual differences in beliefs about police legitimacy. High-legitimacy participants tended to rate lethal police actions following lethal civilian actions as acceptable, but low-legitimacy participants rated lethal police actions much lower. Thus, higher legitimacy seems to yield a willingness to accept lethal police force if civilian resistance is sufficiently severe, while individuals with low legitimacy may never perceive lethal police actions as acceptable. Similarly, low-legitimacy participants were much more accepting of civilian violence against police than high-legitimacy participants, and rated violent civilian reactions as deserving less severe punishment. High-legitimacy participants also tended to rate low-severity police reactions as insufficient when civilian actions were severe, but lowlegitimacy participants did not. Finally, even when actions and re-actions were exactly matched, more severe officer and civilian actions both tended to be perceived as somewhat excessive for all participants regardless of legitimacy beliefs, though likely for different reasons. USE-OF-FORCE APPROPRIATENESS 2

Portions of this work have been previously presented at the Society for Personality and SocialPsychology annual conferences (2017, 2018, 2019) and the associated pre-conferences on Justice andMorality and Law and Psychology. Additionally, separate analyses of these data are reported in a companion manuscript (Celestin & Kruschke, 2019). Although the two manuscripts are based on the same data, the analyses are (a) mathematically different and (b) focus on very different aspects of the models.The latent severity scales reported here remain essentially unchanged when using more elaborate models, so we present the most parsimonious model here in order to maximize the clarity of the latent scale derivation and the implications of those results.

Manuscripts: There are two manuscripts associated with this data set and accompanying analyses. These manuscripts are located in separate subcomponents labeled "Lay evaluations of police and civilian use of force: Action severity scales" and "Lay evaluations of police and civilian use of force: appropriate police reactions are less severe than civilian resistance." Please be sure to download the manuscript of interest from these subcomponents of the main OSF project.