Does discourse breed an appetite for Covid-19 vaccination? An online experiment on group dynamics, arguments, and narratives

我们调查了与疫苗支持者的点对点交流如何影响疫苗怀疑论者接种Sars-Cov-2疫苗的意愿。在许多国家，如德国和美国，对新开发的Sars-Cov-2疫苗的看法在不同的社会人口群体中存在分歧。如果不提高这些国家接种疫苗的意愿，政府将发现，通过接种疫苗实现群体免疫在政治上是昂贵的——如果不是不可能的话。我们进行了一项大规模的调查-聊天实验，将受试者随机匹配到小的同伴小组中，彼此讨论接种Sars-Cov-2疫苗的利弊。我们假设与随机同伴的交流将从两个维度影响接种疫苗的意愿：(1)通过随机同伴群体中的观点分布，(2)通过使用的信息和情绪类型。此外，我们假设意见的可塑性取决于三个因素：(1)考虑到他们的沟通方式，一个主体的接受程度；(2)自己与群体中大多数同伴的意见差异程度；(3)用来支持其主张的同行论证的说服力。大流行严重限制了社会接触，从而严重限制了与同龄人就是否接种疫苗交换意见的机会。相反，人们被锁定在同质的社会环境和/或同性恋的网络回音室中，这降低了信仰的可塑性。我们的实验提供了与随机的同龄人在线见面和讨论的机会。治疗随机分配关于对Covid-19疫苗接种的事前态度的聊天组组成，以及相对于中立对照，启动参与者基于事实或基于叙述的论点类型的干预。通过分析他们在讨论过程中产生的情感和话语动态，我们将阐明在增加或减少接种疫苗意愿方面最有效的论点和说服策略的类型。该领域的在线聊天产生与讨论主题直接相关的富文本数据；而那些有预先形成的观点的人会用论证和叙述来说服别人。因此，在这个项目中，我们将比较基于叙述的论点（即轶事，故事）和基于事实的论点如何在社会群体中改变观点。

We investigate how peer-to-peer communication with vaccination supporters affects the willingness to get vaccinated against Sars-Cov-2 among vaccine skeptics. Opinions on newly developed vaccines against Sars-Cov-2 are split across different socio-demographic groups in many countries, such as Germany and the U.S. Without increasing the willingness to get vaccinated in these countries, governments will find it politically costly – if not impossible - to reach herd immunity by vaccination. We conduct a large-scale survey-chat experiment that will randomly match subjects into small peer groups that chat among each other on the pros and cons of getting vaccinated against Sars-Cov-2. We hypothesize that communication with random peers will affect the willingness to get vaccinated along two dimensions: (1) via the distribution of opinions in random peer groups, and (2) via the types of messages and sentiments used. Furthermore, we hypothesize that opinion malleability depends on three factors: (1) the degree of receptiveness of a subject, given their communication style; (2) the degree of difference between own opinion versus the majority of peers in the group; and (3) the persuasiveness of peer arguments used to back up their claims. The pandemic significantly restricts contacts in society, thus heavily limiting the opportunity to exchange opinions with peers on whether or not to get vaccinated. Instead, people are locked into their homogenous social settings and/or in homophilious online echo chambers, which decreases malleability of beliefs. Our experiment provides the opportunity of meeting and discussing with random peers online. Treatments randomly assign chat-group composition with respect to ex-ante attitude towards Covid-19 vaccination and an intervention priming participants towards either fact-based or narrative-based types of arguments relative to a neutral control. By analyzing the sentiments and discourse dynamics from the text generated during their discussions, we will shed light on the types of arguments and persuasion tactics that are most effective in increasing or decreasing the willingness to get vaccinated. Online chats in the field yield rich text data that directly relate to the discussion topic; and those with pre-formed opinions use arguments as well as narratives to persuade others. Therefore, in this project, we will compare how narrative-based arguments (i.e. anecdotes, stories) and fact-based arguments play a role in shifting opinions across social groups.