5.2.3. Analysis of the Mediation Effect of Information Processing
We test the mediation model again. The results revealed that government trust significantly associated with heuristic processing (b = 0.18, t = 1.79, 95% CI = [-0.02, 0.38], p = .08), supporting Hypothesis 2. However, government trust did not predict system processing (b = 0.06, t = 0.32, 95% CI = [-0.33, 0.46], p = .75); thus, Hypothesis 3 was not supported. When information processing methods and government trust were both entered into the regression equation, heuristic processing significantly negatively associated with risk perception (b = -0.22, t = 2.78, 95% CI = [-0.38, -0.06], p = .006), and the mediating effect of heuristic processing in government trust and risk perception was significant; thus, Hypothesis 4 was supported. Systematic processing did not predict risk perception (b = 0.02, t = 0.44, 95% CI = [-0.06, 0.10], p = .66), and the mediating effect was not significant; thus, Hypothesis 5 was not supported.
GENERAL DISCUSSION
Flood hazards occur very frequently, and how individuals perceive flood hazards is a critical component for formulating risk communication. People’s trust in the government has a great influence on risk perception. In the past, scholars have shared different opinions on whether trust can affect people’s risk perception. Siegrist and Cvetkovich (2000) emphasized that trust has an important influence on risk perception; however, this point has been strongly questioned by Sjöberg (2000). Notably, all prior research was conducted in Western countries.
Our research makes at least three key contributions. First, we identified whether trust affects risk perception. Although, according to cultural theory, different cultural groups worry about distinct issues (Douglas & Wildavsky, 1982; Kahan, Braman, Cohen, Gastil, & Slovic, 2010), risk perceptions vary with cultural biases (i.e., worldviews; Wildavsky & Dake 1990) and experience with previous hazards (Kellens, Zaalberg, Neutens, Vanneuville, & De Maeyer, 2011; Lindell & Hwang 2008; Terpstra 2011). Researchers suggest that considering peoples’ perceptions, social factors, psychological factors, and culture is essential (Bempah & Øyhus, 2017). Study 1 utilized a real, urban waterlogging risk event as the backdrop and examined the relationship between people’s trust in the government and their risk perception. We revealed that the more people believe in the government, the lower their risk perception.
Second, we explored the mechanism between trust and risk perception using flood risk as the backdrop. Studies 2 and 3 used two different methods to jointly test the negative relationship between trust and risk perception using information processing; that is, to answer the question, “How does trust work?”, we clarified the importance of trust in risk perception and elucidated the key influential mechanisms.
Studies 2 and 3 showed that the negative correlation between trust in government and risk perception could be explained by information processing methods. During floods, the higher the public’s trust in the government, the more they will use heuristics to process risk information; their risk perception will also lower. This result is partly consistent with previous results (Trumbo & McComas, 2003, 2008). Trust is a state of mind, which refers to the possibility of accepting vulnerability (Rousseau, Sitkin, Burt, & Camerer,1998) under the premise of making positive expectations of the intentions or actions of others. This mental state of trust makes it possible to reduce the complexity of things and prevent people from thinking too much about potential risks (Siegrist & Cvetkovich, 2000). People tend to use heuristic processing when they have a high level of trust in the government, thus leading to a lower risk perception (Trumbo & McComas, 2008).
Third, this research found that when government trust is low, people have a higher risk perception; however, this relationship cannot be explained by information processing. Contrary to the conclusion drawn by Tortosa-Edo et al. (2014), which was based on chemical pollution risk. Heuristic processing is more dependent on people’s experiences: compared with chemical pollution and nuclear pollution risks, people have more experience with flood risks (Ge et al., 2011); thus, people may have more heuristic processing when facing flood risks. Future research can focus on how different risk events affect people’s psychological processing mechanisms.
This study also has some limitations. First, the sample representativeness was limited. Although adult participants were employed in Study 1, the sampling scope was still limited to young populations, the findings cannot be generalized to older populations. Second, this research utilized self-reported information processing methods, which are prone to key biases (e.g., social approval), this may lead very low reliabilities. Future research needs to explore more effective tools to capture participants’ information processing. Finally, we did not did not measure trust in information but trust in government. Future research can examine how the trust in information influence information processing and risk perception.
CONCLUSIONS
Government trust can influence risk perception of flood, the higher the public’s trust in the government, the more they will use heuristics to process risk information; their risk perception will also lower.