Harvard behavior scientist who studied honesty accused of fabricating data

[MechAg94]

https://nypost.com/2023/06/24/harvard-behavior-scientist-who-studied-honesty-accused-of-fabricating-data/?utm_source=twitter&utm_campaign=nypost&utm_medium=social

That headline is hard to ignore.   

In a blog, called DataColada, run by three behavioral scientists, it alleged fraud in four academic papers that Gino co-authored.

They said they presented evidence of fraud to Harvard in the fall of 2021 tied to a 2012 paper and another three papers she was a part of.

The 2012 paper relied on three separate studies, including one that Gino spearheaded

One of the experiments asked about 100 participants to complete a worksheet of 20 puzzles, and for every puzzle cracked they would get $1, the Times reported.

Participants later submitted a form saying how much money they earned from the puzzle-solving but were led to believe that they wouldn’t get caught if they cheated.
Keep up with today's most important news

Stay up on the very latest with Evening Update.
Enter your email address

By clicking above you agree to the Terms of Use and Privacy Policy.

In reality, researchers would be able to know how many puzzles were solved.

The study determined participants were more truthful if they attested to the accuracy of their responses at the top of the form, and not the bottom, but the three scientists — Uri Simonsohn, Leif Nelson and Joseph Simmons — in their blog post alleged some of the data had been tampered with, according to the Times.

[HankB]

From the brief summary above and in the OP, it's not clear why someone would tamper with the data in this sort of study. Were I the researcher, it really wouldn't matter to me if people were more honest if they attested to their honesty at the top or the bottom of the paperwork.

I will say this sort of "study" doesn't sound like its the kind of work worthy of Ph.D. level researchers - from the summary presented here, it's more on the level of a grammar school science fair project.

[cordex]

Maybe the study had an inconclusive result and did not show a strong signal in either direction. Perhaps the researcher didn’t like the idea that her time had been wasted and thought she could redeem that time investment by tweaking the data so that something publishable was present.

[WLJ]

[230RN]

I'm not surprised.  there's a lot of outright false research and literature generated by academians. 

Publish or perish.

Michael A. Bellesiles
https://en.wikipedia.org/wiki/Arming_America
Fired.

Ward Churchill
https://en.wikipedia.org/wiki/Ward_Churchill
Fired.

These two prominent cases come to mind immediately, but over the years I have established a pretty jaundiced attitude.

I often suspect that data is fudged to satisfy a grantor.  My usual first question when I see "studies show" is "who paid for the 'research?' "

Doubting Thomas was my favorite disciple.

Terry 230RN