The Impacts of Narratives-based Risk Communication on Hazard Preparedness

面对极端事件，当地备灾对于避免灾难至关重要。传统的风险沟通依赖于科学信息来影响备灾，而这种技术信息往往被无效地吸收到人们的风险认知和决策中。防灾在很大程度上是由文化价值观、认知偏差、情感、知识、信息和经验等因素决定的，所有这些因素都是通过人们构建和相互讲述的故事来传达的。这项研究将测试一种创新的基于叙事的风险沟通战略的效果，该战略将科学危险信息定位在当地生产的危险叙述中。通过共同编写的风险说明，有效地将科学信息与个人的风险认知和决策联系起来，可能会提供一种创新的方式来改进灾害准备，这可能会在更广泛的危险问题上体现出来。由于该项目的重点是洪灾事件，这项研究将利用社会行为科学、水文学和计算机科学方面的专业知识。这项跨学科的研究工作侧重于测试共同制作的、基于叙事的风险沟通方法是否比传统的风险沟通方法在改善灾害准备(定义为风险感知和决策)方面更有效。第一个目标是开发联合制作的风险描述，既具有科学准确性，又具有地方相关性。采用以社区为基础的参与性研究(CBPR)方法，将从河流社区收集洪水风险叙述的基线数据，以确定叙述政策框架(NPF)中确定的叙述要素。叙事元素包括叙事结构的子集(例如，人物、情节的使用)和叙事内容(谁扮演英雄、恶棍、受害者的角色)。虽然叙事结构是稳定的，但内容在不同的叙事中有所不同。叙事中的人物及其相关行为在构建不同的现实概念和随后的决定方面起到了形成作用。然后，将使用自然语言处理计算技术从CBPR数据中确定关键的叙述内容，以获得叙述结构和当地语言和图像的叙述内容的最佳组合。随后，研究人员将在每个社区的100年洪水地图中量化、解释和描述与洪水频率分析相关的水文不确定性(数据、模型和自然不确定性)的来源。由此产生的信息将被嵌入到基于CBPR的算法增强的风险描述中，以创建与当地相关的、科学上准确的洪水风险描述。这些风险说明将退回CBPR小组进行调整和验证。第二个目标侧重于通过在这些河流社区的更大人口中进行实验性调查设计，测试这些共同制作的风险叙述作为对灾害准备(即风险感知和预期决定)的叙述处理的效果。不同结构的联合制作叙事(即以英雄为中心的、以受害者为中心的、以英雄和受害者为中心的叙事)将被用作治疗，以测试它们与非叙事性科学声明和不治疗的对照条件相比对危险准备的影响程度。预计这些发现将提供对叙述在传达灾害风险和影响灾害准备方面的力量的洞察。预计研究结果将对参与防灾战略的地方和联邦实体有所帮助。

Local hazard preparedness is vital to avoid disaster in the face of extreme events. Whereas conventional risk communication relies on scientific information to affect hazard preparedness, such technical information is often ineffectively assimilated into people's risk perceptions and decisions. Hazard preparedness is largely shaped by factors such as cultural values, cognitive biases, affect, knowledge, information, and experiences, all of which are communicated through stories that people construct and recount to one another. This research will test the effects of an innovative narrative-based risk communication strategy that locates science hazard information in locally produced hazard narratives. Effectively connecting scientific information to individual risk perceptions and decisions through co-produced risk narratives potentially offers an innovative way to improve hazard preparedness that could translate across hazard issues more broadly. With the project focus on flooding events, this research will draw upon expertise in social behavioral sciences, hydrology, and computer science. This interdisciplinary research effort focuses on testing whether a co-produced, narrative-based risk communication approach is more effective than conventional risk communication at improving hazard preparedness (defined as risk perception and decisions). The first objective is to develop co-produced risk narratives that are both scientifically accurate and locally relevant. Using a community-based participatory research (CBPR) approach, baseline data of flood risk narratives will be collected from river communities to ascertain narrative elements as identified in the Narrative Policy Framework (NPF). Narrative elements include subsets of narrative structure (e.g., use of characters, plot) and narrative content (who is cast in the roles of hero, villain, victim). Whereas narrative structures are stable, content varies across narratives. The characters cast in the narrative and their associated actions are formative in constructing different notions of reality and consequent decisions. Natural Language Processing computational techniques will then be used to identify key narrative content from the CBPR data to obtain the best set of combinations of narrative structure and narrative content situated in local language and images. Subsequently, researchers will quantify, explain, and depict sources of hydrologic uncertainty (data, model, and natural uncertainty) associated with flood frequency analysis in 100-year flood maps for each community. This resulting information will be embedded into the algorithmically enhanced CBPR-based risk narratives to create locally relevant and scientifically accurate flood risk narratives. These risk narratives will be returned to the CBPR groups for adjustment and validation. The second objective focuses on testing the effects of these co-produced risk narratives as narrative treatments on hazard preparedness (i.e., risk perception and intended decisions) with an experimental survey design across the larger population in these river communities. Differently constructed co-produced narratives (i.e., hero-focused, victim-focused, hero & victim focused narratives) will be used as treatments to test the extent to which they influence hazard preparedness in contrast to a non-narrative science statement and a control condition of no treatment. The findings are expected to provide insight into the power of narratives in communicating hazard risk and affecting hazard preparedness. The outcomes are expected to be useful for those local and federal entities involved in hazard preparedness strategies.