Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.


Luis A. Afonso
Posts:
4,758
From:
LIsbon (Portugal)
Registered:
2/16/05


Sufficient large samples lead surely to reject H0 . . .
Posted:
Nov 7, 2012 11:38 AM


Sufficient large samples lead surely to reject H0 . . .
A drawback usually ascribed to NHST: Samples from nature very rarely, or never, are conformable with theoretical model . . . enough to increase sample sizes to be aware. Inference: Worthless to find out if the Null is rejected or not rejected: The result do depend how many observed items used, and worse, one infallibly has to reject it. Analysis__ Two possibilities should be distinguished (if possible): the sample does include extraneous values, and are discarded, or an unknown Population feature come into sight, leading to modify the parameter value we check. Is just, I mean, time to research in depth how to adapt the model to the find. (Why the cygnet is black?) Do remember: *scientific culture* follows the Occam´s razor criterion: The simplest model that is conformable with experience/observation should be followed: en.wikipedia.org/wiki/Occam's_razor
Luis A. Afonso



