Risk Communication to Motivate Flood and Hurricane Risk Mitigation

Developing and Testing Social Norms and Self-Efficacy Messages

JungKyu Rhys Lim
The World Bank

Abstract

At-risk communities can take steps to mitigate hurricane and flood risks, but many choose not to do so. Previous research has not fully examined the role of persuasive messages designed to encourage mitigation measures. This study developed and tested risk communication messages intended to encourage mitigation measures, such as purchasing flood insurance and installing water barriers, using social norms and coping appraisals. Social norms are a social group expectations of how members should behave (e.g. one should protect oneself from flood risk), while coping appraisals are the process in which an individual determines if a behavior is effective and if it can be implemented (e.g., a person thinking about purchasing flood insurance will consider if effectively reduces their risk, as well as if they can afford it). Messages based on social norms and coping appraisals were tested using four online experiments in flood- and hurricane-prone states (N = 5,027). Results indicated that social norms-based messages effectively encouraged at-risk individuals to engage in risk mitigation behaviors. There was no evidence that coping appraisal messages were effective. Theoretical and practical implications are discussed.


Introduction

At-risk individuals can individuals can protect themselves by taking disaster risk mitigation measures, such as installing water barriers and purchasing flood insurance. However, it is challenging to motivate people to recognize risk associated with natural hazards, take mitigation measures, and support related policies. Risk communication strategies may be useful in encouraging such behaviors. However, previous research has not fully explored what types of messages might be most persuasive. For example, rather than developing specific interventions, previous studies have largely focused on exploring and identifying factors that affect individual risk perception and mitigation behaviors (e.g., Lindell & Perry, 20001, 20122; Slotter et al., 2020). To address this gap, researchers have suggested developing and testing communication interventions using experiments (Bamberg et al., 20173; van Valkengoed & Steg, 2019).

The purpose of this study was to develop and test risk communication messages using social norms and coping appraisals to motivate individuals to take specific disaster risk mitigation behaviors. Messages were tested using four between-subject online experiments (2 X 2 X 2 X 2) in flood- and hurricane-prone states (N = 5,027). Only a few studies (Kranzler et al., 20204; Steinmetz et al., 20165) have systematically developed and tested messages based on self-efficacy and social norms, and only a limited number of studies (Howe et al., 20176; Vinnell et al., 20197) have evaluated such messages in disaster mitigation contexts.

Social norms, or subjective norms, are the perceptions of social expectations that can influence the behaviors of individuals (Ajzen, 19918, 20029; Cialdini, 201210). Manipulating social norms has been recommended for persuasive communication (O’Keefe, 200411), including for messages related to disaster risk mitigation (Linnemayr et al., 201612; Meyer & Kunreuther, 201713). Specifically, mitigation behavior can be increased by providing messages for descriptive social norms, injunctive norms that use familiar references (e.g., neighbors, weather forecasters, disaster officials), and disapproval rationale can increase behavioral intentions to engage in risk mitigation. Specifically, descriptive social norms, or norms of ‘is,’ are information about how common behaviors are among group members (Goldstein et al., 200714; Nolan et al., 200815). Conversely, injunctive social norms, or norms of ‘ought,’ are information about whether a social group commonly approves or disapproves of behavior (Cialdini et al., 2012). Finally, disapproval rationale is why people need to engage in behaviors, such as social consequences of not engaging in mitigation behaviors. Thus, this study proposed the following hypothesis:

H1: Descriptive norm messages, injunctive norm messages from weather forecasters, injunctive norm messages using neighbors, and disapproval rationale messages will increase social norms, which in turn, increases mitigation behavioral intentions.

Additionally, capitalizing on coping appraisals, such as those encouraging self-efficacy, may also be a useful approach to encourage risk mitigation behaviors. Self-efficacy-based approaches have also been recommended for use in message designs (O’Keefe, 2004), as self-efficacy explains and predicts disaster risk response behaviors in nonexperimental studies (e.g., Bamberg et al., 2017; van Valkengoed & Steg, 2019).

Based on previous research, there appear to be a few ways to address self-efficacy in risk communication messages. For example, crisis communication scholars recommended providing instructions to help people to physically protect themselves (Coombs, 201916; Sturges, 199417). Additionally, Bandura (200118) argued that self-efficacy can be developed using vicarious experiences by social models (i.e., seeing social models who are similar to the audience perform the behavior), and social persuasion (i.e., encouraging an audience to engage in the behavior). Thus, this study also proposed a second hypothesis:

H2: “Why” (explanation) messages, “how” (instruction) messages, cost (required resource) messages, and “can” (verbal persuasion) messages will increase self-efficacy, which in turn, will increase mitigation behavioral intentions.

Methods

Research Design

This study used four (2 X 2 X 2 X 2) between-subject experiments (N = 5,027) to test the influence of specific types of risk communication messages in motivating risk mitigation behaviors in communities in the United States. Specifically, this study includes four experiments that test (1) social norms messages for purchasing flood insurance (n = 1,252), (2) social norms messages for installing water barriers (n = 1,226), (3) coping appraisal messages for purchasing flood insurance (n = 1,304), and (4) coping appraisal messages for installing water barriers (n = 1,245).

To test social norms messages to encourage at-risk individuals to purchase flood insurance and install water barriers, this study used two 2 (descriptive norm: present vs. absent) X 2 (injunctive norms using weather forecasters: present vs. absent) X 2 (injunctive norms using neighbors: present vs. absent) X 2 (disapproval rationale: present vs. absent) experiments.

To test coping appraisal messages to encourage at-risk individuals to purchase flood insurance and install water barriers, this study used two 2 ("why,” explanation: present vs. absent) X 2 (“how,” instruction: present vs. absent) X 2 (“can,” verbal persuasion: present vs. absent) X 2 (cost, required resource: present vs. absent) experiments with one additional vicarious experience condition with all components present.

This research was reviewed and approved by a university Institutional Review Board (University of Maryland, 1583498).

Participants

At-risk individuals in flood and hurricane-prone states in the United States (N = 5,027) participated in the study between August and September 2020, using Amazon Mechanical Turk (MTurk). This study uses flood and hurricane-prone states in the Gulf Coast and East Coast to increase generalizability and ease of recruitment: Alabama (n = 388), Florida (n = 1,717), Louisiana (n = 300), Mississippi (n = 154), North Carolina (n = 754), and Texas (n = 1,714) (FEMA, 202119).

Participants averaged 21.56 minutes to complete the study. The mean age of participants was 37.68 (SD = 11.57). Regarding gender, 2,470 participants identified as male (49.1%), 2,533 as female (50.4%), and 21 preferred not to identify (0.4%). Participants include Caucasian (n = 2,809), African American (n = 986), Asian (n = 237), Hispanic (n = 569), Native American (n = 327), and those who identified as other or preferred not to say (n = 99). Overall, the participant demographic mirrored the demographics of each state’s residents (U.S. Census, 202120).

Sampling

Participants were recruited through Amazon’s MTurk crowdsourcing platform. Most online data collection, regardless of whether it uses professional panels, student participants, or MTurk, uses convenience sampling and can have potential concerns, such as cheaters, those who take the survey too rapidly , professional survey-takers, or self-selection bias. Still, studies have found that MTurk participants are more representative of the general population, compared to other types of data collection (Berinsky et al., 201221; Kees et al., 201722).

Potential participants were provided a description of the study and completed a brief eligibility screener (e.g., being 18 years or older or living in one of the previously identified states). To ensure data quality, the advertisement for the study was restricted to people in the United States who have MTurk reputations of 95% or higher and have completed at least 100 HITs (i.e., completed 100 tasks on MTurk and did not have more than 5% of their tasks rejected), following previous studies (Cunningham et al. 2017; Peer et al., 2014).


Figure 1. Mock Facebook Posting Used as Experiment Stimuli.

Procedures

Participants were randomly assigned to one of the conditions and exposed to the stimuli. Participants then viewed a mock government campaign (e.g., ready.gov) posting on Facebook (see Figure 1). To ensure exposure to the message, participants listened to professional voiceover audio recordings when reading the post.

Then, participants were asked to respond to measures adapted from previous studies on a 7-point Likert-type scale, such as self-efficacy (Bubeck et al., 2018; Nabi & Myrick, 201923), resource constraints (Bubeck et al., 2018; Poussin et al., 201424), social norms (Bubeck et al., 201325; Vinnell et al., 2019), and behavioral intentions (Nabi & Myrick, 2019; Wilson et al., 201926).

Results

Data Analysis

A structural equation modeling (SEM) with a multiple-indicator-multiple-cause (MIMIC) approach were used to analyze the data (Breitsohl, 201927; Hayes & Preacher28, 2014; Muthen & Muthen, 201729). The model fit was evaluated with Hu and Bentler’s (199930) criteria: RMSEA .06 or lower, SRMR .08 or lower, CFI .95 or higher. The study followed a two-phase modeling process (Anderson & Gerbing, 1988[^Anderson and Gerbing, 1988[), which includes the first measurement phase and the following structural phase. The overall measurement model indicated fit the data well (e.g., chi-square, df, p-value, CFI, RMSEA, SRMR), indicating that the items sufficiently and reliably measured the latent constructs.

Findings

Messages based on social norms and coping appraisals were developed and tested with the goal of encouraging individuals to purchase flood insurance and install water barriers. Specifically, social norms messages included descriptive norms, injunctive norms using weather forecasters, injunctive norms using neighbors, and disapproval rationale (e.g., telling people that they will face social consequences if they do not take mitigation measures) messages. Conversely, coping appraisal messages included explanations for the need to perform behaviors (why), instruction to perform behaviors (how), verbal persuasion (can), required resources (cost), and vicarious experience messages.

Social norms messages showed strong and positive results across the water barrier and flood insurance contexts. In particular, injunctive norms using weather forecasters and disapproval rationale messages increased behavioral intentions across the contexts. However, descriptive norm messages did not increase behavioral intentions (see Figures 2 and 3). Thus, the use of injunctive norms message using weather forecasters (e.g., “All of your local weather forecasters agree that everyone living in hurricane-prone areas should purchase flood insurance”) and disapproval rationale (“Because if you don’t, your damaged home can harm others’ homes and lower your community’s property values”) were supported for both contexts for purchasing flood insurance and installing water barriers. However, injunctive norms message using neighbors (“Most of your neighbors think you should install water barriers”) was only supported for the context for installing water barriers.

Figure 2. Social Norms Messages for Purchasing Flood Insurance: SEM Results


Note: X indicates interaction. * significant at 0.10 level, ** significant at 0.05 level, *** significant at 0.01 level, **** significant at 0.001 level.

Figure 3. Social Norms Messages for Installing Water Barriers: SEM Results


Note: X indicates interaction. * significant at 0.10 level, ** significant at 0.05 level, *** significant at 0.01 level, **** significant at 0.001 level.

Conversely, the coping appraisal messages showing why (explanation), how (instruction), cost (required resource), and can (verbal persuasion) messages—and any of these messages together—decreased behavioral intentions. The vicarious experience message (e.g., “I’m Sam. My partner and I are just like you. Living in [participant’s state] for [years of living that participant entered] years, we knew that hurricanes were a threat”) was much more effective than the "why,” “how,” cost, and “can” messages in a flood insurance context. However, vicarious experience message was not much more effective than no stimuli condition. Therefore, none of the elements of the second hypothesis was not supported.

Discussion

To encourage disaster risk mitigation behaviors, this study developed, tested, and compared effective messages using social norms and coping appraisal (e.g., self-efficacy, response efficacy). No known studies simultaneously have compared these messages’ effectiveness. Social norms messages were more effective than coping appraisal messages.

Social norm messages, such as injunctive norms using weather forecasters and disapproval rationale messages, strongly encouraged behavioral intentions to purchase flood insurance and install water barriers. However, descriptive norm messages did not increase behavioral intentions. For effective risk mitigation communication, organizations can use injunctive norm messages; for example, it is effective to send a message that weather forecasters believe that individuals should take disaster risk mitigation behaviors. Organizations also can use disapproval rationale messages that highlight the social consequences of not engaging in mitigation behaviors.

Conversely, coping appraisal messages did not produce the target behaviors. Most messages about the rationale, instruction, required resources, or verbal persuasion decreased behavioral intentions. Providing information about why, how, and whether at-risk individuals engage in the mitigation behaviors may not have a direct, positive impact. However, vicarious experience messages using a narrative format and matched characters were effective in some mitigation contexts.

Limitations, Strengths, and Future Directions

Like all studies, this study has limitations and strengths. The results cannot be generalized to other disaster types, countries, or cultures. This study used randomized experiments for internal validity. This study relied on previous studies and government communication materials for message wording, visuals, and emojis, to have ecological validity and reflect real-world risk communication. Future studies can examine the impacts of other factors, such as various communication channels, possibly social media, information sources, visuals, and emojis for risk mitigation context (e.g., Buntain & Lim, 2018; Olteanu et al., 201531). Also, participants may have had retrospective biases (Fischhoff et al., 200532; Lim et al., 201933). Lastly, this study developed and tested messages targeting at-risk individuals. Future studies might develop and test messages targeting other stakeholders, such as policymakers.

Conclusions

With increasing climate risks, disaster risk mitigation efforts can save lives. Organizations have spent tremendous amounts of money on disaster relief and recovery after natural hazards resulted in the loss of lives and money. However, investing more in disaster mitigation efforts—including risk communication campaigns that encourage taking mitigation action—could avoid such costs. This study shows that it is important to carefully design messages. Using social norms messages can encourage at-risk individuals to engage in risk mitigation behaviors, while coping appraisal messages can discourage such behavior. The next step is to find better ways of communicating disaster mitigation behaviors to various stakeholders.

References


  1. Lindell, M. K., & Perry, R. W. (2000). Household adjustment to earthquake hazard: A review of research. Environment and Behavior, 32(4), 461-501. https://doi.org/10.1177/00139160021972621 

  2. Lindell, M. K., & Perry, R. W. (2012). The protective action decision model: theoretical modifications and additional evidence. Risk Analysis, 32(4), 616-632. https://doi.org/10.1111/j.1539-6924.2011.01647.x 

  3. Bamberg, S., Masson, T., Brewitt, K., & Nemetschek, N. (2017). Threat, coping and flood prevention–A meta-analysis. Journal of Environmental Psychology, 54, 116-126. https://doi.org/10.1016/j.jenvp.2017.08.001 

  4. Kranzler, E. C., Czajkowski, J., & Chen, L. J. (2020). Identifying Promising Messages to Increase Hurricane Mitigation Among Coastal Homeowners in the United States. Risk Analysis, 40(11), 2313-2328. https://doi.org/10.1111/risa.13560 

  5. Steinmetz, H., Knappstein, M., Ajzen, I., Schmidt, P., & Kabst, R. (2016). How effective are behavior change interventions based on the theory of planned behavior?. Zeitschrift für Psychologie, 224(3), 216-233. https://doi.org/10.1027/2151-2604/a000255 

  6. Howe, P. D. L., Boldero, J., McNeil, I. M., Vargas-Saenz, A., & Handmer, J. (2017). Increasing preparedness for wildfires by informing residents of their community’s social norms. Natural Hazards Review, 19(2), 1-9. https://doi.org/10.1061/(ASCE)NH.1527-6996.0000279 

  7. Vinnell, L. J., Milfont, T. L., & McClure, J. (2019). Do social norms affect support for earthquake-strengthening legislation? Comparing the effects of descriptive and injunctive norms. Environment and Behavior, 51(4), 376-400. https://doi.org/10.1177/0013916517752435 

  8. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179-211. https://doi.org/10.1016/0749-5978(91)90020-T 

  9. Ajzen, I. (2002). Perceived behavioral control, self‐efficacy, locus of control, and the theory of planned behavior. Journal of Applied Social Psychology, 32(4), 665-683. https://doi.org/10.1111/j.1559-1816.2002.tb00236.x 

  10. Cialdini, R. B. (2012). The focus theory of normative conduct. In P. A. M. Van Lange, A. W. Kruglanski, & E. T. Higgins (Eds.), Handbook of Theories of Social Psychology (pp. 295-312). Sage. 

  11. O'Keefe, D. J. (2004). Trends and prospects in persuasion theory and research. In Readings in persuasion, social influence, and compliance gaining (pp. 31-43). Pearson. 

  12. Linnemayr, S., O’Hanlon, C., Uscher-Pines, L., Van Abel, K., & Nelson, C. (2016). Using insights from behavioral economics to strengthen disaster preparedness and response. Disaster Medicine and Public Health Preparedness, 10(5), 768-774. https://doi.org/10.1017/dmp.2016.29 

  13. Meyer, R., & Kunreuther, H. (2017). The Ostrich Paradox: Why We Underprepare for Disasters. Wharton Digital Press. 

  14. Goldstein, N. J., Griskevicius, V., & Cialdini, R. B. (2007). Invoking social norms: A social psychology perspective on improving hotels' linen-reuse programs. Cornell Hotel and Restaurant Administration Quarterly, 48(2), 145-150. https://doi.org/10.1177/0010880407299542 

  15. Nolan, J. M., Schultz, P. W., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V. (2008). Normative social influence is underdetected. Personality and Social Psychology Bulletin, 34(7), 913-923. https://doi.org/10.1177/0146167208316691 

  16. Coombs, W. T. (2019). Ongoing Crisis Communication: Planning, Managing, and Responding. Sage. 

  17. Sturges, D. L. (1994). Communicating through crisis: A strategy for organizational survival. Management Communication Quarterly, 7(3), 297-316. https://doi.org/10.1177/0893318994007003004 

  18. Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52(1), 1-26. https://doi.org/10.1146/annurev.psych.52.1.1 

  19. FEMA. (2021). Data visualization: Summary of disaster declarations and grants. https://www.fema.gov/data-visualization-summary-disaster-declarations-and-grants 

  20. U.S. Census. (2021). State Population by Characteristics: 2010-2019. United States Census Bureau. https://www.census.gov/data/tables/time-series/demo/popest/2010s-state-detail.html 

  21. Berinsky, A. J., Huber, G. A., & Lenz, G. S. (2012). Evaluating online labor markets for experimental research: Amazon.com's Mechanical Turk. Political Analysis, 20(3), 351-368. https://doi.org/10.1093/pan/mpr057 

  22. Kees, J., Berry, C., Burton, S., & Sheehan, K. (2017a). An analysis of data quality: Professional panels, student subject pools, and Amazon's Mechanical Turk. Journal of Advertising, 46(1), 141-155. https://doi.org/10.1080/00913367.2016.1269304 

  23. Nabi, R. L., & Myrick, J. G. (2019). Uplifting fear appeals: Considering the role of hope in fear-based persuasive messages. Health Communication, 34(4), 463-474. https://doi.org/10.1080/10410236.2017.1422847 

  24. Poussin, J. K., Botzen, W. W., & Aerts, J. C. (2014). Factors of influence on flood damage mitigation behaviour by households. Environmental Science & Policy, 40, 69-77. https://doi.org/10.1016/j.envsci.2014.01.013 

  25. Bubeck, P., Botzen, W. J., Kreibich, H., & Aerts, J. C. (2013). Detailed insights into the influence of flood-coping appraisals on mitigation behaviour. Global Environmental Change, 23(5), 1327-1338. https://doi.org/10.1016/j.gloenvcha.2013.05.009 

  26. Wilson, R. S., Zwickle, A., & Walpole, H. (2019). Developing a broadly applicable measure of risk perception. Risk Analysis, 39(4), 777-791. https://doi.org/10.1111/risa.13207 

  27. Breitsohl, H. (2019). Beyond ANOVA: An introduction to structural equation models for experimental designs. Organizational Research Methods, 22(3), 649-677. https://doi.org/10.1177/1094428118754988 

  28. Hayes, A. F., & Preacher, K. J. (2014). Statistical mediation analysis with a multicategorical independent variable. British Journal of Mathematical and Statistical Psychology, 67(3), 451-470. https://doi.org/10.1111/bmsp.12028 

  29. Muthén, L. K., & Muthén, B. O. (2017). Mplus User’s Guide. Eighth Edition. Muthén & Muthén. https://www.statmodel.com/download/usersguide/MplusUserGuideVer_8.pdf 

  30. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1-55. https://doi.org/10.1080/10705519909540118 

  31. Olteanu, A., Vieweg, S., & Castillo, C. (2015). What to expect when the unexpected happens: Social media communications across crises. Proceedings of the 18th ACM Conference on Computer Supported Cooperative Work & Social Computing, 994–1009. https://doi.org/10.1145/2675133.2675242 

  32. Fischhoff, B., Gonzalez, R.M., Small, D.A., & Lerner, J.S. (2005). Evolving judgments of terror risks: Foresight, hindsight, and emotion. Journal of Experimental Psychology: Applied, 11, 124-139. https://doi.org/10.1037/1076-898X.11.2.124 

  33. Lim, J. R., Liu, B. F., & Egnoto, M. (2019). Cry wolf effect? Evaluating the impact of false alarms on public responses to tornado alerts in the southeastern United States. Weather, Climate, and Society, 11(3), 549-563. https://doi.org/10.1175/WCAS-D-18-0080.1 

Suggested Citation:

Lim, J. (2021). Risk Communication to Motivate Flood and Hurricane Risk Mitigation: Developing and Testing Social Norms and Self-Efficacy Messages (Natural Hazards Center Mitigation Matters Research Report Series, Report 5). Natural Hazards Center, University of Colorado Boulder. https://hazards.colorado.edu/mitigation-matters-report/risk-communication-to-motivate-flood-and-hurricane-risk-mitigation

Lim, J. (2021). Risk Communication to Motivate Flood and Hurricane Risk Mitigation: Developing and Testing Social Norms and Self-Efficacy Messages (Natural Hazards Center Mitigation Matters Research Report Series, Report 5). Natural Hazards Center, University of Colorado Boulder. https://hazards.colorado.edu/mitigation-matters-report/risk-communication-to-motivate-flood-and-hurricane-risk-mitigation