3 ms·
Wrong is hardly the ideal metric here. Being less wrong than random can beat being correct and utterly trivial.
by Retric 4y ago
Wrong is hardly the ideal metric here. Being less wrong than random can beat being correct and utterly trivial.
- mgamache 4y agoThe standard is a p value of .9 or .95. Not a flip of the coin.
- Retric 4y agoThat’s hardly the only way to be wrong. Suppose you made a mistake in the synthesis process and the dosage described doesn’t match the dosage revived by patients. Now the paper is incorrect but possibly still quite useful.