Rethinking the Role of Affect in Risk Judgment: What We Have Learned From COVID-19 During the First Week of Quarantine in Italy.

Due to COVID-19 spreading in Italy, on March 11 the Prime Minister of Italy declared a lockdown and imposed severe restrictive measures impacting citizens' freedom at several levels. People were required to stay at home and go out only to satisfy basic needs. Several risk models have postulated a link among online searching behavior, affect, anxiety, and complaints by individuals toward government restrictions (GR), which emerged as also related to an increased perception of knowledge toward risk. However, to date, no study has addressed how these key risk-related aspects (i.e., affect, anxiety, perceived knowledge on risk, and risk dimensions) can act jointly to orient online health information-seeking behavior, and people's complaints toward GR imposed during the lockdown. This study investigated the mechanisms underlying online health information-seeking behavior and people's complaints toward the government's restrictions during a COVID-19 emergency in the Italian population. Drawing from the health belief model (HBM), which postulates a link between sociodemographic variables, risk, and affect dimensions in emergency, we assumed risk factors as predictors of affect and anxiety, which, in turn, were posited as mediators between risk dimensions, online health information-seeking behavior, and complaints toward GR. Participants (1,031) were involved during the first week of the quarantine (March 11-18) and completed an online survey composed of (i) an adapted version of the Italian Risk Perception Questionnaire; (ii) the Italian Positive (PA) and Negative Affect (NA) Schedule (PANAS-10); (iii) the State Anxiety Scale (STAI-Y1); (iv) personal knowledge measure about novel coronavirus; (v) item measuring information search behavior regarding the novel coronavirus; (vi) measure of the complains regarding GR; and (vii) sociodemographic questions. General linear models and structural equation modeling (SEM) were carried out to test the model. Sociodemographic and cognitive factors predicted the participants' affect and anxiety, which, in turn, motivated and fully mediated both information search behavior and complaint toward GR. This research can offer useful suggestions for policy-makers during the COVID-19 emergency, and it advanced the knowledge on the risk-emotion link in emergency situations.