Re:Let's hear scientists with different Covid-19 views, not ...
by magicalhippo on Monday April 27, 2020 @11:22PM
(#23002999)
> You can't "disprove" data.
In the interest of correctness, yes you can.
To disprove means "to prove to be false or wrong"[1], and that's exactly what the OPERA collaboration did to their own data that showed super-luminal neutrinos, when they later discovered it was caused by a bad fiber connection[2].
There's also the case of BICEP2[3], where the raw data was not in dispute, but rather the inevitable post-processing required to extract the "science signal" from the raw data.
This is of course why scientists prefer to have multiple independent measurements of the same phenomena, to eliminate possible sources of error and bias. The LHC for example has two independent main detectors[4], built by different teams using different approaches, for this very reason.
That said, at least in the physical sciences (I don't follow life sciences as closely), cases like OPERA and BICEP2 are exceptions.
In the interest of correctness, yes you can.
To disprove means "to prove to be false or wrong"[1], and that's exactly what the OPERA collaboration did to their own data that showed super-luminal neutrinos, when they later discovered it was caused by a bad fiber connection[2].
There's also the case of BICEP2[3], where the raw data was not in dispute, but rather the inevitable post-processing required to extract the "science signal" from the raw data.
This is of course why scientists prefer to have multiple independent measurements of the same phenomena, to eliminate possible sources of error and bias. The LHC for example has two independent main detectors[4], built by different teams using different approaches, for this very reason.
That said, at least in the physical sciences (I don't follow life sciences as closely), cases like OPERA and BICEP2 are exceptions.
[1]: https://www.merriam-webster.com/dictionary/disprove
[2]: https://en.wikipedia.org/wiki/Faster-than-light_neutrino_ano...
[3]: https://en.wikipedia.org/wiki/BICEP_and_Keck_Array#BICEP2
[4]: https://home.cern/science/experiments