According to Thomas Kuhn, scientific revolutions always take place in response to anomalous data: When data are presented that do not concur with an established paradigm, the paradigm will shift. We believe that such a paradigm shift is taking place with respect to our quantitative research practices as we speak. The series of reported anomalous data leading up to this shift started in neighboring disciplines, such as medicine (Ioannidis, 2005) and psychology. It may have started with a widely discussed article published in The New Yorker by Jonah Lehrer (2010), who observed that, mysteriously, "the truth wears off' (para. 1). Only a year later, Daryl Bem's (2011) nine-experiment proof for precognition provided another anomaly, this time published in the prestigious Journal of Personality and Social Psychology: A result that hardly anyone believed, but that met the same methodological standards as any other social science article. Almost at the same time, Simmons, Nelson, and Simonsohn (2011) demonstrated the power of "p-hacking." They showed how a little instrumental tweaking of collected data--everything within generally accepted boundaries of good conduct--produced ludicrous findings: People become one-and-a-half years younger after listening to a Beatles song. Anomalous data, again.
展开▼
机译：根据托马斯·库恩（Thomas Kuhn）的观点，科学革命总是针对异常数据而发生：当提供的数据与既定范式不一致时，范式将发生变化。我们相信，就我们所说的定量研究实践而言，这种范式正在发生。导致这一转变的一系列报告的异常数据始于邻近学科，例如医学（Ioannidis，2005年）和心理学。它可能始于乔纳·莱勒（Jonah Lehrer）在《纽约客》（The New Yorker）（2010）上发表的广泛讨论的文章，该文章神秘地观察到“真相消失”（第1段）。仅一年后，达里尔·贝姆（Daryl Bem）（2011）的九-用于预知的实验证据提供了另一个异常，这一次发表在享有盛名的《人格与社会心理学杂志》上：结果几乎没人相信，但达到了与任何其他社会科学文章相同的方法论标准。 ，Nelson和Simonsohn（2011）证明了“ p-hacking”的威力。他们展示了如何对收集到的数据进行一些工具性的调整（一切都在公认的良好行为范围内），产生了可笑的发现：人们变得一劳永逸听了甲壳虫乐队的歌曲后，年轻了半岁，再次出现了异常数据。
展开▼
