In the 2014 article “Know Thy Avatar: The Unintended Effect of Virtual-Self Representation on Behavior,” Yoon and Vargas report that playing a video game for 5 min as one of three avatars had large effects on participants’ subsequent aggressive and prosocial behavior, as measured by the amount of chocolate chips or hot chili sauce they poured for another participant to eat. These results have considerable theoretical importance for psychologists’ understanding of the effects of video games on behavior, the nature of self-representation, and the causes of aggression. However, there are two concerns regarding the accuracy of the data and, thus, the validity of the inferences. First, the effect size is improbably large. Second, certain cells of the reported data contain remarkably little variance. I present data that shows such little variance is improbable. Together, these issues suggest that there may be some potentially invalidating error in data collection or analysis.

中文翻译：

---

对 Yoon 和 Vargas (2014) 的评论：来自难以置信的不变数据的难以置信的巨大影响

在 2014 年的文章“了解你的化身：虚拟自我表征对行为的意外影响”中，Yoon 和 Vargas 报告说，作为三个化身之一玩电子游戏 5 分钟对参与者随后的攻击性和亲社会行为有很大影响，根据他们倒给另一名参与者吃的巧克力片或辣椒酱的量来衡量。这些结果对于心理学家理解电子游戏对行为的影响、自我表现的本质和攻击性的原因具有相当大的理论意义。然而，关于数据的准确性以及推论的有效性存在两个问题。首先，效应量大得不可思议。其次，报告数据的某些单元格包含的差异非常小。我提供的数据表明，这种差异很小是不可能的。总之，这些问题表明在数据收集或分析中可能存在一些潜在的无效错误。