Question: Can you think of a widely accepted scientific claim that is based on a single study and no replications? [Operationalize "widely accepted" as routinely in the discipline's core knowledge textbooks.] If so, what is the claim and what was the study?
-
Show this thread
-
Replying to @BrianNosek
Theoretical physics is often based on zero studies (directly), yet ends up in the discipline's core textbooks. Thousands and thousands of examples. Eg the Higgs Boson and gravity waves and BEC were in textbooks for decades before confirmation.
3 replies 1 retweet 33 likes -
Replying to @michael_nielsen
Good point. I should narrow the question to account for theory-based claims. But, in your examples, would experts agree that there are no studies offering supporting evidence, or that the critical experiments haven't been done yet, or something else?
2 replies 0 retweets 5 likes -
Replying to @BrianNosek
Something like, say, gravity waves is tough to fit into this narrative. Of course, GR had supporting evidence. And Bell et al's pulsar results provided indirect results (in the 1970s, IIRC, though). But it's really not difficult to imagine a world where LIGO found nothing.
1 reply 0 retweets 3 likes -
Replying to @michael_nielsen @BrianNosek
In general these situations are very interesting. Instruments sometimes cost hundreds of millions of dollars. So replication is, in the short term, very difficult.
1 reply 0 retweets 4 likes -
Replying to @michael_nielsen @BrianNosek
I wrote a short note about some related ideas here: http://michaelnielsen.org/blog/science-beyond-individual-understanding/ …
1 reply 0 retweets 5 likes -
Replying to @michael_nielsen @BrianNosek
Even when replication is done using a different instrument, there's still the potential for problems. Eg data processing may be done using the same linear algebra libraries (eg BLAS). Those libraries certainly have bugs. What if the bugs cause an artifact mistaken for a signal?
1 reply 0 retweets 2 likes -
Replying to @michael_nielsen @BrianNosek
In any given instance this is unlikely. Over the long term I have no doubt at all this kind of thing is going to cause some very embarrassing mistakes.
1 reply 0 retweets 1 like
Something about your proposed narrowing makes me uncomfortable. In disciplines with very strong theory you'd _expect_ to be more comfortable with single confirmations.
-
-
Replying to @michael_nielsen @BrianNosek
Eventually this shades into engineering. When Lockheed designed the first stealth fighter they were in some sense doing an experiment. But they really did know what the radar profile would be, in advance, because the theory at hand was so strong.
0 replies 0 retweets 1 likeThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.