The abundance of research concerning the level of flood risk perception makes it a dynamic field of empirical research, the resources of which should interest decision makers managing flood risk. The strong diversity of the collected empirical work (e.g., research methods and areas of research) provides a broad view of the applied research approaches.
Rationalists focus on the quantitative approach, probability, and consequences, and constructivists are more interested in the qualitative dimension. The constructivist approach is criticized by researchers for its qualitative character and difficulty in operationalization.
The topics and research interests concerning factors or groups of factors that may influence the perception of flood risk varied over time. In the beginning, attention was paid to the rationalistic approach (mainly to direct experience, socioeconomic and demographic profiles, and information, and later also to residence characteristics and individual's physical location). Over time, the amount of constructivist research increased, particularly that related to indirect experience and a political context. Research interests and their variations over time could result from the availability of data and development of research tools. For example, it is much easier to obtain (and measure) metric information on respondents (e.g., age, gender, education) than their more intimate (personal) data, such as religion and relationships with family or neighbors, and that on historical and cultural backgrounds. With the development of geographical information systems, the scope, detail, and availability of data increased, including techniques for delimiting flood zones, flood characteristics, identification of an individual's physical location, and residence characteristics.
The choice of factors for the perception of flood risk is often simply a matter of interest of an individual researcher. It may also be determined by the research results of other literature due to the easiness of the later discussion on the results obtained. This reasoning can sometimes become a mechanism of a "self-winding circle," resulting in the excessive expansion of one sphere of research at the expense of another.
Contemporary empirical works on flood risk perception are dominated by a rationalistic approach that psychometrically searches for cognitive models of behaviors when there is flood risk; however, this creates a heterogeneous field of research in terms of basic theoretical issues, methodology, and more importantly, the results obtained. Often, statistically obtained dependencies are mutually exclusive. Physical, cognitive, behavioral, and demographic–economic variables alone cannot adequately explain attitudes toward flood risk. One limitation of some rationalistic studies is the "flattening" of variables or the inability to consider them because of external factors. For example, in the case of a location variable, there was a high degree of homogeneity, which resulted in lowering the importance of this variable in the results. The exclusion of women from the study of Muslim communities resulted from cultural and religious factors.
Thus, studies on perception in the constructivist approach are at an early stage of development. Currently, they are only a supplement to rationalistic research and not a mainstream tendency in the field of flood risk perception. However, the results of studies of flood risk assessment by the public in a social, political, cultural, religious, or historical context provide a clear picture.
The analysis of the collected empirical works also indicates mutual relations between individual factors of a rationalistic and constructivist nature. Most often, the influence of rationalistic factors results from constructivist conditions. For example, media coverage contributes to the perception of risk mainly when a recipient has no personal direct experience of the flood; other people's experiences (e.g., the effects of the flood experienced by friends, relatives, neighbors, and other people) have a weaker impact on the perception of risk. Trust in the authorities becomes more important when there is little individual knowledge of risk, and trust in flood protection measures (confidence in their effectiveness) often results from the previous experience of flooding. Furthermore, gender plays a greater role in shaping the perception of flood risk in countries where legal and cultural differences between the sexes are more visible.
Moreover, within the group of constructivist factors, some relations can be observed. Information provided at the local level (from neighbors, family, and friends) is more important in developing the risk perception than the institutional impact is. The political and economic system makes the role of the media in society different (greater control of the authorities results in a more limited amount information on risks). Different historical backgrounds influence the power of social bonds and the type of social capital, shaping the role of indirect experiences in flood risk perception.
The results of this study show a slow shift toward research on soft flood risk perception factors, conducted in the constructivist approach. This is because the role of soft factors in shaping flood risk perception is increasingly important. The greater population mobility causes the mixing of cultures and religions and modification (loosening) of social ties. In the era of a digital information society, the role of media is increasingly important.
Some constructivist factors, such as the influence of media or neighbors, friends, and family on flood risk perception, are more controllable than rationalistic factors (e.g., direct experience of flooding, or gender or age). Constructive studies therefore provide more useful information for improving flood risk management.
Notably, the development of civilization has changed the sources of information about threats, which should be reflected in the research on risk perception. The level of knowledge of risk should not be determined solely based on the knowledge of flood hazard maps, but also, and perhaps primarily, on the basis of media coverage.
Therefore, in the practice of flood risk management, particular attention should be paid to the role of social media in shaping attitudes toward risk. According to the availability heuristic, people remember the hazards that are more dramatically and spectacularly presented better than those that have occurred. In the era of a digital society, electronic media might be more effective in communicating risk. Neighborly ties, by contrast, are loosening up and will be of minor importance in determining the role of indirect experience in shaping attitudes toward risk. It is worth noting the potential of using artificial intelligence (AI) to study the perception of flood risk based on social data. Since it is possible to profile people based on their social network entries, it is also possible to find some common characteristics for groups of people who are more resistant or vulnerable to flood risk based on such networks.
The important information on flood risk management that can be derived from this work is that the influence of rationalistic factors on the level of flood risk perception often results from the constructivist context. Thus, knowing specific local political, cultural, historical, religious, and social conditions makes it possible to more effectively moderate the attitudes of the inhabitants toward flood risks.
The importance of the context in the research on the perception of flood risk has slowly begun to be recognized in the reviews. For example, Raska (2015) perceives the causes of differences in the perception of environmental hazards (e.g., flood risk) occurring between residents of Western Europe and Central and Eastern Europe in different political and cultural contexts (historically conditioned). The post-communist political system (once ensuring a strong position of central government bodies) of the countries of Central and Eastern Europe shape the current passive attitude of the communities of these countries toward floods. In societies of Central and Eastern Europe, decentralized and liberalized as a result of systemic transformation, prevailing materialistic orientation causes greater anxiety and awareness of environmental threats than for the citizens of Western European countries with a post-materialist orientation. Moreover, Boholm (1998) reviews the results of international research draws attention to their diversity, which may result from different geographical, national, and cultural contexts of the studied countries.
In conclusion, constructivist research points to the moderating role of context in studies of flood risk perception. Constructivist factors weaken or strengthen the impact of rationalistic factors on the level of flood risk perception in general and on its three aspects: preparedness, fear, and awareness. A constructivist approach to studying flood risk perception can deepen the understanding of the various, often antagonistic, results of rationalistic research.
The research on flood risk perception should be reoriented. Because of the low compliance of the results of rationalistic research, manifested by the unclear influence of physical, cognitive, behavioral, and demographic–economic factors on flood risk perception, the focus of flood risk research should be shifted to studying the role of contextual factors in shaping risk perception. It is necessary to assess how the public interprets flood risks socially and culturally, and not merely in terms of metric data. The study of risk perception is a social study; thus, the specifics of the studied society (its worldview, culture, history) should be considered. According to ethnographers, human social existence is culturally variable. Perceptions of events and phenomena are conditioned by values that differ locally. Brikholtz et al. (2014) express the same opinion and agree with the need to comprehensively reinvigorate flood risk research, supported by a more constructive approach to flood risk management. Tierney (1999) argues for the use of a constructivist approach by saying that it provides mechanisms by which specific risk perception can be disseminated and legitimized at a wider social level.
However, complete disregard for the rationalist approach is not possible in studies of flood risk perception. Notably, the constructivist and rationalist perspectives are insufficiently reliable and credible. As Renn (1989) aptly points out, they are unable to provide solid guidance on their own, and public values, rational decision-making, and scientific knowledge should be reconciled through a “well-managed discourse.”
In connection with the above, there is a need to revive the study of flood risk perception toward a more holistic view, namely, the development of a constructivist approach to provide a more comprehensive understanding of the factors affecting the perception of flood risk. The constructivist paradigm therefore recommends a more comprehensive analysis of how the socio-cultural context shapes a broader understanding of risk. Research teams should include geographers and scientists representing other scientific disciplines, such as psychology, sociology, cultural studies, religious studies, ethnography, political science, history, social communication, and media.
The postulated direction of flood risk research is part of the research concepts of vulnerability, capacity, and resilience. It closely refers to the innovative approach to developing a social resilience model, which would be based, among other things, on the perception of flood risk (Bradford et al, 2012). The concept of resilience assumes a positive adaptation of society to the existing developmental threats. A resilient society is one that adapts continuously and flexibly to changing circumstances. Community resilience has been highlighted as one of four priority mechanisms for disaster risk reduction worldwide (Schelfaut et al., 2011). The role of risk perception in improving the resilience of people and communities is widely recognized as an important part of the wider area of risk research (Burns, Slovic, 2012). Risk perception is often viewed as a key element of vulnerability assessment.
In many studies, the perception of risk is considered a factor of vulnerability (e.g., Messner and Meyer, 2006; Kuhlicke et al., 2011), but we can agree with Brikholtz et al. (2014) that they rarely elaborate on the perception of flood risk in a substantive manner. Some resilience studies often consider the important role of risk perception as an argument for better risk communication in flood management strategies (e.g., Schelfaut et al., 2011). However, based on the literature on flood risk management, the perception of Kuhlicke and Steinführer (2013) is that the concept of resilience is largely in its infancy, and we agree. At present, relatively few empirical studies have examined the perception of flood risk in terms of resilience. The statement in Brikholtz et al. (2014, p. 15) that the concept of resilience "should, we argue, be used to greater effect in underpinning flood risk research" exactly corresponds to the recommendations made in this work regarding a renewed research agenda for flood risk perception.