Academia is BROKEN! - Harvard Fake Data Scandal Explained

This week a top Professor at Harvard University was exposed for data FRAUD. The evidence is damning, and it is hard to see how Francesca Gino can argue her way out of it. This looks bad for Gino, but also it looks bad for behavioural science in general. She isn't the only example of data fraud in the industry either, so if you want me to cover more of this type of content, let me know in the comments below!

The Articles/Blog Posts:

Thank you to Uri, Joe and Leif. You guys are the real MVP's.

Follow me:
Behavioral Science Instagram: @petejudoofficial
Instagram: @petejudo
Twitter: @petejudo
LinkedIn: Peter Judodihardjo

[thewolf]

Having published in a handful of scientific/academic journals, I have not once been asked for actual deidentified data. The expectation is that researchers conduct themselves with honesty and integrity, but obviously that isn't universally practiced.

[GiovannaIwishyou]

Maybe she wouldn't fake the data if there was a pledge of honesty at the begining of her paper.

[SnapScienceOfficial]

The fact that this was only caught because she was so terrible at faking the data really raises eyebrows. How much data is faked by competent fraudsters?

[samuelmingo]

I was doing a meta-analysis once for a bio-medical project. I came across a research paper that I was pulling data from. The papers contents read good and showed a positive correlation in the outcomes of the intervention. I went to looking at the tables to pull the hard data and I just couldn't make sense of what I was reading. It felt very off. I'm not great at understanding the deep depths of statistics, so I took that paper to a professor of statistics, who had to take a moment after looking at the tables. As it turned out, the tables had been carefully put together to obscure that there wasn't much of a correlation at all. The authors didn't blatantly lie in the paper, but the text implied one thing, while the tables obscured the truth of another thing. It literally took a PhD in statistics to see what they were doing.

[woodificould]

One of the reasons I dropped out of my PhD program. The "publish or perish" motto has legitimately ruined academia

[GhostOfSnuffles]

The correlation between ethics professors and unethical behavior is staggering.

[nukestrom]

Academia is broken because of the behavior of journals. Decades ago, negative results were publishable as it helped fellow researchers not to follow the same path or learn what doesn't work, but nowadays they are thrown out by journals because journals demand only positive results. So, this practice of manipulation is becoming increasingly common.

[grig]

The most surprising thing about this is how bad they are at faking their results

[chrisS]

More people should consider that finding out that your hypothesis didn't pan out is a good thing too. You still made progress by figuring out which route you should avoid taking.

[joechip]

The researchers who challenged her findings aren't vigilantes, they're professors at universities. Challenging your colleagues's work is meant to be part of the job. It's kind of like calling the SIU vigilantes.

[darrenmiliband]

I used to work in a lab, and after putting my blood and my sweat into research, I generated excellent data. One day the professor showed up and claimed that he needed to get a patent for his company and with my work but without my name being acknowledged and asked for my work. Immediately after he received the work, he told me that he would give me another project to work on as he wanted to show investors that he was the one who generated the data in order to secure more funding. when I protested his decision, he fired me and stole my notebook. As an international student, I had no way of getting compensated and was left in limbo. People like these should be removed from academia!

[jabir]

It's actually hilarious that all of these papers are about honesty

[evethel]

There is a saying where I leave; “don’t believe any statistic, that you didn’t fake yourself”

[WillyOrca]

She should just rush and publish a paper titled "How reliable are Academic Studies? Determining the likihood of falsified data passing peer review."

[Sparts]

Man, it’s almost like she was given an incentive from an invested source to lie about her performance. Someone should do a study about that.

[Hellooo]

We need to force journals to publish statistically insignificant results. The pressure to publish <0.05 is absurd and is a big part of problems like this.

[richardplane]

Before we get into whether she was fudging the results or not, the fact that she was trying to find a causal link between test result honesty and cleansing products tells you all you need to know about her and the institution.

To that end, I'd like to apply for her now-vacant role with funding for the following hypotheses:
1. Can I attract better looking birds driving a Ferrari or a beat up Chevy?
2. Do Coldplay play better gigs when I have a front seat, VIP ticket, compared to when I don't ?
3. The statistical probability of a coin toss outcome conducted in the Bronx compared to a four star hotel in Bali.

If the Provost can forward the application papers ill be right over .

[Matthyk]

What is shocking to me is how unsophisticated the fake data is.

[douginorlando]

One correction … this proves getting CAUGHT being a fraud is unacceptable in the profession BUT being a fraud is how to rise to the top in the profession

[ellenchavez]

The thing about research is that it's not hard to get the answer you want: that's a simple matter of switching statistical operations until you get the answer you want.

The trick to research is asking the right questions; developing the right tools.

A better question is why people are dishonest to begin with.