@jessesingal .05 is arbitrary anyway, so you see papers from time to time with .1 marked as significant
Are there normal circumstances under which a researcher would say p<.1 is cool rather than p<.05? Could I be missing simple explanation?
-
-
-
@wccubbison Gotcha. Do you normally need to explain why you're loosening the definition? - Show replies
New conversation -
-
-
@jessesingal absolutely. Could set p-values even higher in some cases, way lower in others. It's about degree of control over circumstances. -
@jessesingal basically the more you determine constraints, lower you need to set p-value. Naturalistic experimentation can use much higher. - Show replies
New conversation -
-
-
@jessesingal Journals often have rules. Sometimes you can report <.1 if what you really mean is a directional hypothesis <.05 (same thing)Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@jessesingal Is p-hacking a "normal circumstance"? - End of conversation
New conversation -
-
-
.
@jessesingal Yes, missing sthg. Problem with p-values is: ppl rely to much on them. Only trust if people replicate their finding in ++ expeThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@GabrielRossman@jessesingal I never believed the argument p-value should change with sample size. Some samples are bigger. Life isn't fair.Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@GabrielRossman@jessesingal bigger issue is the number of tests. Check out the boss p-values in genome-wide studies http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2865141/ …Thanks. 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.