Social media users are not equally susceptible to all misinformation. We call ``diverse misinformation'' the complex relationships between human biases and demographics represented in misinformation, and its impact on our susceptibility to misinformation is currently unknown. To investigate how users' biases impact susceptibility, we explore computer-generated videos called deepfakes as a type of diverse misinformation. We chose deepfakes as a case study for three reasons: 1.) their classification as misinformation is more objective; 2.) we can control the demographics of the persona presented; and 3.) deepfakes are a real-world concern with associated harms that need to be better understood. Our paper presents a survey (N=2,000) where U.S.-based participants are exposed to videos and asked questions about their attributes, not knowing they might be a deepfake. Our analysis investigates the extent to which different users are duped and by what perceived demographics of deepfake personas. First, if users not explicitly looking for deepfakes are not particularly accurate classifiers. Importantly, accuracy varies significantly by demographics, and participants are generally better at classifying videos that match them (especially male, white, and young participants). We extrapolate from these results to understand the population-level impacts of these biases using an idealized mathematical model of the interplay between diverse misinformation and crowd correction. Our model suggests that a diverse set of contacts might provide ``herd correction'' where friends can protect each other's blind spots. Altogether, human biases and the attributes of misinformation matter greatly, but having a diverse social group may help reduce susceptibility to misinformation.
翻译:社交媒体的用户并不同样容易受到所有错误信息的影响。 我们称之为“ 错误错误信息 ”, 人类偏见和人口结构之间的复杂关系, 错误信息代表着复杂的人类偏见和人口关系, 其影响目前还不得而知。 为了调查用户偏见的影响, 我们探索计算机生成的视频“ 深假” 是一种不同的错误信息。 我们选择了深假作为案例研究, 原因有三:1) 他们被归类为错误信息更客观; 2. 我们可以控制所介绍的人的人口结构; 和 3. 深刻假信息是真实世界对相关伤害的关切,需要更好地了解。 我们的论文展示了调查( N=2 000), 美国的参与者在调查中接触到视频, 并询问有关其属性的问题, 我们不知道它们可能是一个深刻的错误信息。 我们的分析调查了不同用户被贬低的程度, 以及人们所认识的深假人的人口结构。 首先, 如果用户没有明确寻找深假数据分类, 并不是特别准确的易变数据; 确实, 准确性数据差异差异很大, 但是, 参与者们通常会更清楚地区分视频,, 。