Intellectual Humility & the importance of knowing you might be wrong

[GordMay]

We’ve seen some very high-profile examples, lately, of how overconfident leadership can lead to ruinous outcomes.

Julia Rohrer (et al) wants to create a radical new culture for social scientists. A personality psychologist at the Max Planck Institute for Human Development, Rohrer is trying to get her peers to publicly, willingly admit it when they are wrong.

To do this, she, along with some colleagues, started up something called the Loss of Confidence Project. It’s designed to be an academic safe space for researchers to declare for all to see that they no longer believe in the accuracy of one of their previous findings. The effort recently yielded a paper that includes six admissions of no confidence. And it’s accepting submissions until January 31.

Loss-of-Confidence Project ➥ https://lossofconfidence.com/

Putting the Self in Self-Correction
“Scientific self-correction is often construed as an outcome of the activities of the community as a whole. In contrast, cases in which researchers publicly point out errors in their own studies are rare and deemed unusual. Here, we argue that such individual self-corrections would be beneficial for the scientific community. In an online project, we invited researchers to submit statements describing how they have lost confidence in a finding they had previously published...”
More ➥ https://psyarxiv.com/exmb2

[moctrams]

This is off topic. Keep politics out of this forum.

[LakeSuperior]

Darn, I wanted to see all the climate change research papers that would be posted.

[atoll]

the earth really is flat,members from all around the world can agree on this.......

[SV THIRD DAY]

Originally Posted by LakeSuperior

Darn, I wanted to see all the climate change research papers that would be posted.

Ha ha ha....
If I said that I'd get a visit to chat room jail again for being offensive....

[Lake-Effect]

Originally Posted by SV THIRD DAY

Ha ha ha....
If I said that I'd get a visit to chat room jail again for being offensive....

You wouldn't have said just that though.

[AndyEss]

Originally Posted by SV THIRD DAY

Ha ha ha....
If I said that I'd get a visit to chat room jail again for being offensive....

You are only offensive when you keep on insisting on eating street fish tacos

[fish53]

Shoot, I can't post on here anymore.

[Mike OReilly]

Interesting Gord. I see this as part of the quest for the social “sciences” to become real sciences. So much of social science (sociology, psychology economics, anthropology, etc…) is pretty loosey-goosey with regard to scientific rigour.

Actually, I just finished reading Seth Stephens-Davidowitz’s book “Everybody Lies”. His main thesis, other than the fact that we’re all liars, is that Big Data is finally allowing social science to become legitimate sciences. Big data allows for rigorous experimentation and analysis over legitimate sample groups. This is as opposed to most social science experiments which usually involve a small group of university students.

[gamayun]

Read the link. This is specifically related to the field of psychology, which has been going through its own catharsis recently after finding that many of their studies could not be replicated. Transparent data and reproducibility are keys to good science; understanding science is key to a smart society. We're getting there on the former; we ain't there yet on the latter.

[john61ct]

Yes this is not a political topic.

Respect for proper, verified consensus science is critical to our shared reality.

Proper evidence based science must inform all policy decisions, not delusions nor corrupt self interest.

And a properly scientific attitude must remain humble and open to new data and interpretations.

Much science wrt **human** behaviour,
is not settled, in need of constant re-testing and verification.

[Stu Jackson]

Originally Posted by john61ct

Yes this is not a political topic.

Respect for proper, verified consensus science is critical to our shared reality.

Proper evidence based science must inform all policy decisions, not delusions nor corrupt self interest.

And a properly scientific attitude must remain humble and open to new data and interpretations.

Much science wrt **human** behaviour,
is not settled, in need of constant re-testing and verification.

Very succinctly presented. Thanks.

[GordMay]

Good Science is Hard - Publishing Isn't (Dodgy Journals publish papers in exchange for a fee)
Sometimes there's error; and sometimes outright fraud.


The International Journal of Advanced Computer Technology recently accepted for publication a paper titled “Get Me Off Your ****ing Mailing List”, whose text was nothing more than those seven words, repeated over and over for 10 pages. The paper was not actually published as the authors' declined to pay the required US$150 article processing charge.
➥ http://www.scs.stanford.edu/~dm/home/papers/remove.pdf

Two other journals allowed an engineer (Alex Smolyanitsky) posing as Maggie Simpson and Edna Krabappel to publish a paper, “Fuzzy, Homogeneous Configurations.” The text of the Simpson, et al. paper wasn’t even coherent. Smolyanitsky just used a bunch of jargon from the text generator SCIgen.
The Journal of Computational Intelligence and Electronic Systems accepted the paper immediately, while the other waited a month before publishing - and they now keep sending Smolyanitsky invoices for a $459 publishing fee.
➥ https://cdn2.vox-cdn.com/uploads/cho..._Simpson.0.pdf

[gamayun]

Originally Posted by GordMay

The text of the Simpson, et al. paper wasn’t even coherent. Smolyanitsky just used a bunch of jargon from the text generator SCIgen.

Hmmm. It looked coherent to me. Sounds exactly how my techie son talks....

An issue with scientific journals (the legitimate ones - not those that are just making money to publish complete crap for people who need to pay to get cited work) is that they have been shown to be biased to publish only statistically significant results. No effects findings are also informative, but are being shut out. The great thing about science though is that it is constantly being challenged to ensure we get it right and this will make the studies much better over time, although there will always be people who game the system. As we are beginning to understand more about the non-rational decisions that humans make -- all the time -- we will also see ways to counteract these tendencies. Perhaps even better is when AI takes over and separates us from bad choices based on our biases and emotions

[Eigenvector]

For a most excellent read on this subject and many others I'd suggest:


The Skeptics Guide to the Universe


I just finished it and am going to start over. Just too much good insights that I didn't retain the first time through.