We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
cover of episode Why Is There So Much Fraud in Academia? (Update)

Why Is There So Much Fraud in Academia? (Update)

2024/12/26
logo of podcast Freakonomics Radio

Freakonomics Radio

AI Deep Dive AI Insights AI Chapters Transcript
People
B
Brian Nosek
诺斯克是开放科学和可复制性研究的领先人物,通过开放科学中心和其他倡议推动科学研究的透明度和完整性。
D
Dan Ariely
F
Francesca Gino
L
Leif Nelson
M
Max Bazerman
S
Simine Vazire
S
Stephen Dubner
以《怪诞经济学》系列著名的美国作家、记者和广播电视人物。
U
Uri Simonsohn
Topics
Stephen Dubner:本集探讨学术欺诈的普遍性及其对社会的影响,并以Francesca Gino和Dan Ariely的案例为例进行分析。学术欺诈不仅损害了学术界的声誉,也对公共政策和社会生活造成了负面影响。 Brian Nosek:学术研究的完整性至关重要,因为其结果会影响到公共政策和社会生活等诸多方面。现有的学术奖励制度鼓励发表论文,而非注重研究的透明性和严谨性,这导致了数据造假等问题的出现。学术界的奖励制度与理想的科学目标之间存在冲突,这使得研究人员面临在职业发展和科学道德之间的两难选择。 Leif Nelson, Uri Simonsohn, Joseph Simmons:数据可乐达团队通过对数据的统计分析和模式识别,可以识别出学术研究中的欺诈行为。P值操纵(p-hacking)等行为可能会导致不可靠的研究结果,即使研究人员在大多数情况下遵循了规则。一些看似无害的行为,例如剔除异常值或变换数据,可能会导致欺诈行为的发生。 Max Bazerman:在“签名在前”论文的研究过程中,Max Bazerman 对数据的收集过程参与较少,对数据的有效性产生了怀疑,并进行了多次重复实验。Max Bazerman 认为自己有义务纠正学术记录中的错误,并对自身在学术研究中对数据的验证不足感到内疚。 Dan Ariely:Dan Ariely承认2012年研究论文中的数据存在问题,但他声称自己对数据操纵行为不知情。 Simine Vazire:学术界缺乏公开批评的文化,这使得学术欺诈行为难以被发现和纠正。人性的自我欺骗能力很强,研究人员可能出于多种动机而夸大或歪曲研究结果。现有的学术奖励制度鼓励学术欺诈行为。心理学领域正经历一场危机,这场危机涉及到研究的完整性、可信度和科学方法等多个方面。

Deep Dive

Key Insights

Why did Dan Ariely and Francesca Gino's research on signing at the top of forms claim such a high average mileage?

The average mileage in the study was reported between 24,000 and 27,000 miles per year, which is unusually high compared to the average American driver who drives around 13,000 miles a year. Ariely initially claimed the drivers were senior citizens in Florida, but the data provided by the insurance company showed no difference between those who signed at the top and those who signed at the bottom.

Why did Max Bazerman decide not to take the Data Colada evidence to Harvard initially?

Bazerman felt overwhelmed by the evidence and acknowledged his involvement as a co-author on multiple papers with Francesca Gino. He was concerned about the potential personal and professional consequences and felt it was not his place to be the central figure in exposing the fraud.

Why did the Data Colada team start investigating academic research papers for fraud?

The Data Colada team, consisting of Leif Nelson, Uri Simonsohn, and Joseph Simmons, started investigating academic research papers because they noticed a pattern of findings that didn't align with their intuition and statistical impossibilities in the data. They aimed to demonstrate how easy it is to produce significant but unreliable results through practices like p-hacking.

Why is academic fraud a significant concern beyond just the academic community?

Academic fraud is a significant concern beyond the academic community because it can lead to public policies based on false or unreliable research, causing real-world harm. For example, the Wakefield scandal on vaccines and autism has had a corrosive impact on public health. Additionally, it can lead to wasted resources and undermine public trust in academic research.

Why did the original 'signing at the top' paper attract so much attention and implementation?

The 'signing at the top' paper attracted significant attention and implementation because it presented a simple and easily implementable solution to reduce dishonesty. The effect was reported to be large and significant across multiple studies, and it was endorsed by prominent researchers like Dan Ariely and Francesca Gino. Many firms and government agencies, including Lemonade Insurance and the U.S. government, adopted the practice based on these findings.

Why did Max Bazerman, a senior researcher, not interact with the original data in the 'signing at the top' study?

Bazerman trusted his junior colleagues, particularly Francesca Gino, to handle the data collection and analysis. As a senior researcher, he focused on other aspects such as funding, administrative issues, and mentoring young scholars, which meant he did not closely examine the data himself.

Why did the Data Colada team find the insurance data in the 'signing at the top' study suspicious?

The Data Colada team found the insurance data suspicious because the histogram of miles driven by the study subjects showed a nearly uniform distribution from zero to 50,000 miles, which is not what real data typically look like. Real data usually form a bell curve with most people clustered around the average. The insurance company later confirmed that the data Ariely used were significantly different from what they provided.

Why does Simine Vazire believe that the field of psychology is in crisis?

Simine Vazire believes psychology is in crisis due to a lack of integrity and credibility. The field is undergoing intense self-examination, and while many researchers are committed to scientific values, the prevalence of unreliable findings and the lack of robust replication efforts have raised concerns about the overall quality of research.

Why did Max Bazerman and his team fail to replicate the 'signing at the top' effect?

Bazerman and his team failed to replicate the 'signing at the top' effect multiple times, even after adjusting their methodology. When they conducted a large-scale replication using more than 10 times the original sample size, they still found no effect, suggesting the original findings were not robust.

Why did the Data Colada team use a blog to share their findings?

The Data Colada team used a blog to share their findings because it allowed them to communicate more quickly and directly than the traditional journal review process. They wanted to provide short, accessible critiques and updates on their investigations without waiting years for peer review.

Shownotes Transcript

Some of the biggest names in behavioral science stand accused of faking their results. Last year, an astonishing 10,000 research papers were retracted. In a series originally published in early 2024, we talk to whistleblowers, reformers, and a co-author who got caught up in the chaos. (Part 1 of 2)