Why Bad Science Spreads

Follow-Up Interviews On My Second Channel

Many reasons have been given for the replication crisis and while they have some truth to them, I think a simpler theory is in order. Giraffes are to blame. Let me explain.

Research & Footnotes

Follow

Music
Martin Fowler - Ships I
Anthony Barilla - Fallout 1, Fallout 2a
Aaron Sprinkle - Feet Down
licensed on soundstripe
b-side . - leaving EP

[CoffeeBreaks]

Incidentally, this video is also subject to natural selection. Youtube is a system of subject to variation, (different types of content) heritability (shares, other people discussing topic, etc) and selection. (Watch time/Youtube algorithm). Thanks for helping it spread!

[Touj-jpdt]

The Goodharts Law reminds me of the Cobra Effect - that during the British Empire years in india, the British tried to reduce the number of snakes in Delhi by paying locals a bounty to catch snakes. The locals simply bred snakes and sold then to the British.

[sebastianalvarez]

As a researcher I'm honestly having this problem right now. Being rushed constantly to work on finishing papers. Even though most of the research demonstrates that most studies lack consistency and accuracy, and that we need more longitudinal studies. Sadly if you're not constantly publishing, you're not being "efficient" and "productive"

[schodes]

Great ending from Dr. Smaldino. "Show me the incentive and I will show you the outcome" -Munger

[ingGS]

In my PhD program we are required to publish AT LEAST three articles to graduate. This is insane since most of the data and end of experiments happen around year 3, so you have to write them within 1 to 2 years, plus take courses. How good can these papers be? One of the most cited works on my field took 10 years to complete, the author has already passed, it was common back then to take all this time to replicate the experiment, analyze the data, etc. now we are in a rush to publish more and quickly.

[cannaroe]

I was 4.5 years into my PhD with 1 month to go, just writing up the last chapter of my thesis, when my supervisor realised my PhD was about how a small bug in a data format (not code, the data) results in big problems for all of Genomics. My stipend was cut and i became homeless in another country. The bug still exists to this day, and to my knowledge no one actively working in bioinformatics is talking about it. Perhaps because they know if they did, they would also be removed from the discourse

[khreanos]

This also applies outside of science. Like Modern politics. And other jobs I see all of this.

[joel]

These videos drop so much knowledge. YouTube is doing itself a disservice by not promoting this channel more. Keep up the great work CB.

[asdasfghgf]

In my school's astronomy department it is a well worn joke that anything published in Nature and Science (the most impactful journals in the field) are always wrong.

[Mylada]

During the review of my last article one reviewer commented "nice job", nothing else.

[highly_elusive]

When I suggested to my professor to do a replication study: "but that's not new right? Journals won't want to publish that"
Also in my (limited) experience many researchers, especially in the biomedical field, don't have a very good grasp of statistics. They'll focus on the clinical part they are experts in, and when conducting studies themselves, will use only basic, standard tests that may not be a good match for the data. There's needs to be more cooperation between statisticians and clinical researchers.

[baileymoody]

One thing that you couldve mentioned, is the legal authority that government has to redact scientific findings for public safety.

[Vits]

As a huge nerd and a science person myself, I've seen way too much of this over the years.

This breaks my heart.

[MHConverse]

Oh my gosh THIS VIDEO. I've always struggled with explaining to people why I've always been nervous to go into science as a full blown career even though I love it. It draws attention to a topic that needs to be talked about desperately. Thank you so much for bringing it to light. :)

[luilu]

Great video. I've heard in the academic field by some of my professors that, a lot of them will publish the fluff and work on their passion piece in the in between time. Much like an office worker who finishes his work for the day in 5 hours, and does whatever to improve theirselves.

[MedlifeCrisis]

The footage superimposed onto the laptop at 03:51 is oddly specific. It is from a Soviet film about Vladimir Demikhov's work on transplanting dogs' heads. Is that a callback to another video where you've covered that before? I ask as I was working on a video about Demikhov around the time you were making this video. How odd.

[jdk]

Reminds me of journalism today. It isn't about truth whatsoever, it's about generating web traffic, dollars, and brand loyalty.

[Pyriphlegeton]

Let's just make it part of every scientists education to try to replicate and improve one paper.

That might actually be a great training method to learn about potential flaws.

[HumansOfVR]

*_Why did the giraffe get bad grades?_*








He had his head in the clouds.

[HistoryDose]

From police interrogations to science itself, I can always count on your videos changing the way I think about something! As a content creator, that's really inspiring