Individuals tend to follow choices and behaviors that are common among others, indicating majority influence. Nevertheless, majority behaviors that appear to be consistent can be generated by different factors during the decision-making process; hence, the current study addressed whether people consider the source of majority behavior and follow the majority differently when that behavior is formed under different choice conditions. The participants were asked to select a safe exit after observing others' choices in simulated fire escape scenes, and their probability of adopting the majority choice was tested when all or some of the other persons were free or constrained to choose an exit for escaping under different sizes of majority. The results showed that people are more likely to follow the majority's choice when faced with available options than when constrained to make a common choice, and an increase in majority size leads to an increase in majority influence only when the majority behavior is formed under free choice. Importantly, when the majority behavior is generated by a computer algorithm, the majority influence is not affected by the availability of options, ruling out nonsocial differences as contributors to source-dependent majority influence. These findings indicate that people do not simply engage in any form of majority behavior at face value but rather distinguish between different types of majority behavior while following the majority. This finding highlights the fact that people account for rational factors and seek to maximize information gains when following the majority.

中文翻译：

---

明智地跟随他人：多数人的影响力取决于多数人的行为是如何形成的


个人倾向于遵循其他人中常见的选择和行为，这表明大多数人具有影响力。然而，在决策过程中，看似一致的大多数行为可能是由不同因素产生的；因此，当前的研究探讨了当这种行为是在不同的选择条件下形成时，人们是否会考虑多数行为的根源并以不同的方式追随多数。要求参与者在模拟火灾逃生场景中观察其他人的选择后选择安全出口，并在其他全部或部分人自由或被迫选择不同条件下逃生的情况下，测试他们采取多数选择的概率。大多数的大小。结果表明，人们在面临可用选项时比被迫做出共同选择时更有可能遵循多数人的选择，并且只有当多数人行为是在自由选择下形成时，多数人规模的增加才会导致多数人影响力的增加。重要的是，当多数行为由计算机算法生成时，多数影响力不受选项可用性的影响，从而排除了非社会差异作为来源依赖多数影响力的贡献者。这些发现表明，人们并不是简单地从表面上参与任何形式的多数行为，而是在追随多数的同时区分不同类型的多数行为。这一发现凸显了这样一个事实：人们在追随多数时会考虑理性因素并寻求最大化信息收益。