In science, theoretical claims are cheap. Evidence is hard. Even highly plausible claims based on a strong nomological net of indirect evidence can fall apart when well-designed studies directly test a claim’s predictions. In research on intergroup bias, for example, people may report ideological or prejudicial biases but fail to discriminate anyway due to systemic or psychological forces that constrain the influence of bias (e.g., LaPierre, 1934; Massey, 2015; Wicker, 1969). Clark and Winegard (this issue) theorize that political biases in the predominantly liberal social sciences are motivated by a collection of sacred values called equalitarianism, which is comprised of the assumptions that: groups do not differ genetically on socially valued traits; bigotry is ubiquitous and the only cause of group differences; people who assert biological group differences are justifying their own prejudice; and groups can and should be made equal within society. After, they review five sources of evidence that support their theoretical claim. In this commentary, I assess the strength of five sources of evidence that Clark and Winegard describe to support their claim. They first reported (1) strong evidence for selfreported equalitarian or anti-conservative biases among social scientists. If social scientists say that they will discriminate, then we should believe that they will if given the opportunity and ability to do so. However, this source of evidence may not generalize to bias within social science in situ. Institutional safeguards and mitigating social-psychological factors may prevent the expression of ideological biases within social science. To directly examine equalitarian biases in the actual behavior of social scientists, the authors describe (2) experimental evidence demonstrating equalitarian biases among social scientists, (3) weak/non-effects that became popular because they were amenable to equalitarianism, (4) studies that ignore hypotheses that violate equalitarianism, and (5) studies that frame results to avoid opposing equalitarianism. My review finds these four sources of evidence to be indirect at best, with supporting evidence that is anecdotal, cherry-picked, selectively interpreted, or plainly irrelevant to the core theoretical claim. The lack of strong direct evidence for the authors’ claims should give readers pause. In science, it is often said that “extraordinary claims require extraordinary evidence.” At the same time, ordinary claims require ordinary evidence. Weak or anecdotal evidence is not an adequate substitute for the quantitative standards that social scientists hold themselves to. Indirect evidence of ideological bias is not sufficient, however plausible that evidence may be. Making the ordinary claim that equalitarian biases pervade the social sciences still requires ordinary evidence demonstrating that equalitarian biases pervade the social sciences.

中文翻译：

---

普通主张需要普通证据：社会科学中缺乏对平等主义偏见的直接支持

在科学中，理论主张是廉价的。证据很难。当精心设计的研究直接测试声明的预测时，即使是基于强大的间接证据法学网络的高度可信的声明也可能崩溃。例如，在对群体间偏见的研究中，人们可能会报告意识形态或偏见偏见，但由于限制偏见影响的系统或心理力量而无法进行歧视（例如，LaPierre，1934；Massey，2015；Wicker，1969）。克拉克和温加德（本期）的理论是，在占主导地位的自由社会科学中，政治偏见是由一系列称为平等主义的神圣价值观所驱动的，它由以下假设组成： 群体在社会价值特征上没有遗传差异；偏见无处不在，是群体差异的唯一原因；主张生物群体差异的人是在为自己的偏见辩护；群体可以而且应该在社会中平等。之后，他们审查了支持他们的理论主张的五个证据来源。在这篇评论中，我评估了克拉克和温加德描述的五个证据来源的强度，以支持他们的主张。他们首先报告了（1）社会科学家自我报告的平等主义或反保守偏见的有力证据。如果社会科学家说他们会歧视，那么我们应该相信，如果有机会和能力，他们会这样做。然而，这种证据来源可能无法推广到原位社会科学中的偏见。制度保障和减轻社会心理因素可能会阻止意识形态偏见在社会科学中的表达。为了直接检查社会科学家实际行为中的平等主义偏见，作者描述了 (2) 证明社会科学家之间存在平等主义偏见的实验证据，(3) 由于适合平等主义而变得流行的弱/非影响，(4) 研究忽略违反平等主义的假设，以及 (5) 研究框架结果以避免反对平等主义。我的评论发现这四个证据来源充其量是间接的，支持证据是轶事、精心挑选、选择性解释或与核心理论主张明显无关。作者的主张缺乏强有力的直接证据应该让读者停下来。在科学中，人们常说“非凡的主张需要非凡的证据”。同时，普通索赔需要普通证据。微弱或轶事证据不足以替代社会科学家坚持的定量标准。意识形态偏见的间接证据是不够的，无论证据多么可信。提出平等主义偏见普遍存在于社会科学的一般主张仍然需要普通证据证明平等主义偏见普遍存在于社会科学中。