Roberts and colleagues focus on two aspects of racial inequality in psychological research, namely an alleged underrepresentation of racial minorities and the effects attributed to this state of affairs. My comment focuses only on one aspect, namely the assumed consequences of the lack of diversity in subject populations. Representativeness of samples is essential in survey research or applied research that examines whether a particular intervention will work for a particular population. Representativeness or diversity is not necessary in theory-testing research, where we attempt to establish laws of causality. Because theories typically apply to all of humanity, all members of humanity (even American undergraduates) are suitable for assessing the validity of theoretical hypotheses. Admittedly, the assumption that a theory applies to all of humanity is also a hypothesis that can be tested. However, to test it, we need theoretical hypotheses about specific moderating variables. Supporting a theory with a racially diverse sample does not make conclusions more valid than support from a nondiverse sample. In fact, cause-effect conclusions based on a diverse sample might not be valid for any member of that sample.

中文翻译：

---

理论检验研究中受试者需要多样性的神话：罗伯茨等人对“心理学研究中的种族不平等”的评论。 (2020)


罗伯茨及其同事在心理学研究中关注种族不平等的两个方面，即所谓的少数族裔代表性不足以及这种状况造成的影响。我的评论仅集中在一个方面，即主体群体缺乏多样性的假设后果。样本的代表性在调查研究或应用研究中至关重要，这些研究或应用研究检查特定干预措施是否适用于特定人群。在我们试图建立因果律的理论检验研究中，代表性或多样性并不是必需的。由于理论通常适用于全人类，所有人类成员（甚至美国本科生）都适合评估理论假设的有效性。诚然，一种理论适用于全人类的假设也是一个可以被检验的假设。然而，为了测试它，我们需要关于特定调节变量的理论假设。用种族多样化样本支持理论并不比非多元化样本的支持更有效。事实上，基于不同样本的因果结论可能对该样本的任何成员都无效。