Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Lagslayer]

So apparently an unknown amount of accepted science might be unrepeatable or exaggerated. Especially recent things.

Basically statistical anomalies look cool and get attention, and publishing is biased towards results that confirm the original study. People with null results find it difficult to get their papers published because they're boring and there's not any money to be made from them. This means lots of negative results are effectively hidden.

Eventually ideas get established and it becomes cool to disprove accepted scientific truth, but tons of damage can be done in the meantime. Especially when this affects medicine.

Spoiler:  Obligatory xkcd comic. (click to show/hide)

[Sergarr]

Not that surprising for medicine imo, considering the greed of pharmaceutical companies.

At least 1/X of the research people will just go out and say 'hey bossdude, it plain doesn't work'.

And then that person gets fired for "spreading panic and misinformation".

[scrdest]

Not that surprising for medicine imo, considering the greed of pharmaceutical companies.

At least 1/X of the research people will just go out and say 'hey bossdude, it plain doesn't work'.

And then that person gets fired for "spreading panic and misinformation".

And then the person doing the firing gets fired for wasting company money on having to hire a new researcher in the place for a guy trying to save the company money.

[Bauglir]

So, either there needs to be further work on reigning in psychological biases even beyond what double-blind studies offer, or else the Imprisoning God* is fucking with us.

*I'm linking directly into the middle of a multi-chapter story, so spoilers ahead.

[Akura]

Not that surprising for medicine imo, considering the greed of pharmaceutical companies.

At least 1/X of the research people will just go out and say 'hey bossdude, it plain doesn't work'.

And then that person gets fired for "spreading panic and misinformation".

And then the person doing the firing gets fired for wasting company money on having to hire a new researcher in the place for a guy trying to save the company money.

And thus creates the spiral of people getting fired for wasting money, resulting in a constant spread of panic and misinformation.

[scrdest]

No. You got rid of an incompetent, unprofessional employee who fired a competent employee for silly reasons. This is a pretty good reason to fire someone.

[Frumple]

Bad PR and/or inefficiently monetized product are also good reasons to fire someone. Just because a drug doesn't work, doesn't mean you can't make money from it.*

And gods help the poor moral bastard that actually says something outside the company, because the industry damn sure won't.

*There is an absolutely massive non-FDA approved market, hint hint. Also we totally do get recalls and whatnot -- just like with mining companies, it's irrelevant if you have to stop and/or eat fines if you still end up profiting before you get shut down. Better a minor profit and easily buried/obfuscated bad PR (Who actually reads the FDA's recall list? How many times do such things actually make the news?) than no profit and sunk R&D costs.

And given that some drugs take decades before the side effects start kicking in...

[Il Palazzo]

So apparently an unknown amount of accepted science might be unrepeatable or exaggerated. Especially recent things.

Basically statistical anomalies look cool and get attention, and publishing is biased towards results that confirm the original study. People with null results find it difficult to get their papers published because they're boring and there's not any money to be made from them. This means lots of negative results are effectively hidden.

Eventually ideas get established and it becomes cool to disprove accepted scientific truth, but tons of damage can be done in the meantime. Especially when this affects medicine.

Here's a counterpart to that:
http://neuro.plos.org/2014/10/13/this-weeks-most-discussed-plos-neuroscience-article-the-influence-of-partner-specific-memory-associations-on-picture-naming-a-failure-to-replicate/

Considering our current scientific environment, in which the most novel, positive findings are lauded, many researchers might hesitate to report a failed self-replication for fear of interfering with their research trajectory or compromising their reputation. However, Brown-Schmidt and Horton (2014) has served as an exemplar of transparency in scientific reporting, and the authors’ open sharing of their null findings has been received with overwhelming positivity from the scientific community.

The positive response from the community indicates that it's become aware of, and is on a right way of self-correcting the bias.
So not all is doom and gloom in the land of science.

[Sheb]

Certainly not, science is still, by far, the best intellectual tool we ever designed. (Sorry philosophers) However, we need to be aware of those bias, and find ways to improve.

[olemars]

Replication of scientific results should be mandatory for every student on their way to a master's degree or similar. It's perfect as project work.

[Levi]

Two neat things I saw in the news today.

Lockheed Martin is hoping to have fusion power going in about 10 years.

Proof of concept quantum AI.

[Sirus]

Proof of concept quantum AI.

If telling the difference between handwritten characters counts as "AI", then someone should examine the local ATMs which, in my experience, have never failed to read handwritten numbers since they were introduced a couple years back :/

[Frumple]

Certainly not, science is still, by far, the best intellectual tool we ever designed. (Sorry philosophers)

What? Why are you apologizing to the people that originally designed science?

[Sergarr]

Proof of concept quantum AI.

Quantum image recognition device, you mean.

We've had image recognition devices since 1950s.

Also, did China just outdid the USA in science, or is this my imagination?

[Levi]

Proof of concept quantum AI.

Quantum image recognition device, you mean.

We've had image recognition devices since 1950s.

Also, did China just outdid the USA in science, or is this my imagination?

Image recognition counts as AI.  Or at least it did when I was getting my degree.