It’s like ‘Simon says’ but with ‘complete and sufficient’
Does a statistical analysis, informed by valid assumptions, always capture all of the evidence contained in numerical data, or is it POSSIBLE to have numerical evidence that is not detectable via statistics?
#epitwitter #datascience #machinelearning
-
-
-
Not sure I follow. Can you elaborate?
-
Sorry, busy day... the OP said ‘valid’ and then the Q asked about something very close to the definition of statistical sufficiency
-
Fair enough.
End of conversation
New conversation -
-
-
What a sneaky question! Love it.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Is 'evidence' a trick word?
-
Actually I answered to the first bit. Stopped at OR...
End of conversation
New conversation -
-
-
In an overly pedantic sense, absolutely. Conditioning on one model may not have power to say much (ie fitting all random effects), but there is certainly signal that can be captured by remodeling (ie marginalized variance components)
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
So, respondents are to answer the second leg — the possible part, yes?Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
and twitter users (sample) are quite different from non-twitter users.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
As they say, all models are wrong, some are useful. Our assumptions are also based on some beliefs that may not mesh well with reality
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Do graphs count as statistics? Mechanistic models parameterized with stats?
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.