Should I Stay Or Should I Go - Online Experiments On Self-Protection Under Threatening Situations

Online survey experiments (6) were designed using the PsyToolkit platform and the 3D visualization capabilities of Pathfinder and auditory and visual stimuli to represent threatening situations. Each participant faces three trials (fire, explosion and shootings) under different conditions: physical context (either enclosure or open area) and social context (with or without virtual confederates). Decisions (fire: leave/stay; explosion: leave/helping others; shootings: run/hide) and the time taken for such decisions (with time pressure of 10s) were the measured dependent variables. Following testing, a questionnaire was conducted to collect sociodemographic characteristics and impressions of participants.

Results: A pilot was conducted involving 41 participants (24 female, M(age) = 41.9 years; SD = 11.3) performing 2 out of the 6 designed tests, allowing us the possibility to know whether the designed experiment fulfilled the purpose of the study. The preliminary results presented here will be extended with responses from experiments currently being conducted (>1,000 participants) that we will include in the submitted version of the paper.

Preliminary findings: The proportion of people who decided to leave was lower than other options e.g. 74% of participants choose help others in case of explosion. We found no clear associations between sociodemographic variables and decisions. This may be due to the limited pilot sample. Previous training was associated with the decision to leave in all threats (χ2 = 5.62, p = .017). The association between the actions of the confederates (social influence) and the self-protective decisions were significant (χ2=11.95, p <.001). Yet, differences in decision times were found to be significant across trials (Kruskal-Wallis, H(2) = 6.8, p = .033).