4. Discussion

Our research shows that disinhibition and meanness, considered as dispositional individual differences, are related to various forms of support for radical group actions. We also have preliminary support for the role of a group's definitions and norms. That is, people with tem- peramental predispositions could be more prone to support violent actions on behalf of the groups that normalise violence. This fits well with the present understanding of radicalisation. For example, if the reason for radicalisation is a disproportionate commitment to ends that is linked to the devaluation or suppression of alternative (i.e., norma- tive, peaceful) considerations, then people who more easily devaluate alternatives (i.e., because of a lack of consideration of others' well- being) should be more prone to radicalisation. Similarly, people for whom radical behaviours are not only a means to highly valued ends but are also intrinsically reinforcing (i.e., these radical behaviours sa- tisfy the need for thrill-seeking) should be more prone to radicalisation. Here we see the important role of individual differences. Meanness was repeatedly (in three independent samples) related to acceptance of violent social change. This is aligned with clinical and personality research as meanness is understood as empowerment through cruelty and deficient empathy (Patrick et al., 2009). With such disdain for close attachments with others and engagement in active exploitativeness and confrontation, the positive attitudes towards vio- lence could be strengthened by the lack of considerations about the possible human cost of this violence and defiance of authority. Interplay between group identification and disinhibition has been noted as especially relevant for understanding willingness to engage in radical actions on behalf of the right-wing group. Previous studies pointed to the group norms and values as factors linked to the decision to use radical means. Studies show that group radicalisation and po- larisation are largely determined by important group norms, and these norms are often responsible for uninhibited behaviour (if that is a part of a prevailing group norm). In the context of CA, it is possible that impulsive individuals with hostile tendencies will collaborate with other people in a group, especially when group norms justify the violent means chosen by them. In this sense, radical collective actions might be a way to achieve personal benefits (e.g., allowing them to explicitly express a lack of restraint or disregard for others, namely out-group members). The interplay between group norms and individual differ- ences should be explored in future research. CRediT authorship contribution statement Tomasz Besta: Conceptualization, Methodology, Formal analysis, Data curation, Writing - original draft, Project administration. Beata Pastwa-Wojciechowska: Conceptualization, Methodology, Data cura- tion, Writing - review & editing. Michał Jaśkiewicz:Conceptualization, Methodology, Writing - review & editing. Andrzej Piotrowski: Conceptualization, Data curation, Writing - review & editing. Marcin Szulc: Conceptualization, Data curation, Writing - re- view & editing.

Appendix A. Details on participants, measures and procedure

One item attention check was included in the questionnaire to examine if participants follow the instruction: ‘This is an attention-checking question. We are checking whether people filling out the survey do not answer at random. Please, mark the number 2 below’. We included in the final analyses only people who do not failed the attention-checking question (877). Of those, some of the participants failed to answer items from specific scales, thus number of participants in each analysis may vary. Table 1 includes descriptive statistics. for the variables used in research (and number of participants that answered items from each measure).