Drexel dragonThe Math ForumDonate to the Math Forum



Search All of the Math Forum:

Views expressed in these public forums are not endorsed by Drexel University or The Math Forum.


Math Forum » Discussions » sci.math.* » sci.stat.math.independent

Topic: No-Null HST . . .
Replies: 9   Last Post: Feb 20, 2013 3:31 AM

Advanced Search

Back to Topic List Back to Topic List Jump to Tree View Jump to Tree View   Messages: [ Previous | Next ]
Luis A. Afonso

Posts: 4,615
From: LIsbon (Portugal)
Registered: 2/16/05
Re: No-Null HST . . .
Posted: Feb 18, 2013 12:39 PM
  Click to see the message monospaced in plain text Plain Text   Click to reply to this topic Reply

There are situations where all intents to be cool, calm and polite is impossible . . . The event is the jeremiad a lot of Psychologists (and other people) against the NHST (null hypotheses significance tests) used to, seemly restlessly, to exercise.
This time we are dealing with ?How Do We Do Hypotheses Testing? Author: J. Gill.
artsci.wustl.edu/~jgill/papers/hypos.pdf

At the paper´s final conclusive 3 section

Not sufficiently educated think that:

__1__The NHST´s intends to make sure the Null Hypotheses H0 is true (or untrue). NOT AT ALL: they simply measure its plausibility through the p-value: it is completely impossible to get certainly from events/procedures ruled by chance.
In this context p-value has an immutable meaning: by our (arbitrary) criterion to separate, by alpha, plausible from intolerable risk not to reject H0: if we get p-value < alpha do suggests (the word was carefully chosen) that H0 is probably untrue. It worth to note that, under H0 true, the set of p-values are uniformly distributed in [0, 1].
Other current criticism is that p-value does not inform how large is the difference between the observed value of the parameter and the one from which the null is unlike to be taken as true.
This point revels/suggests a serious fail in what concerns statistical tests Psychologists education. The general formula for the test of a parameter X can be written:
_______ V = (X - (X0 + d))/ s
By X0 is expressed the value relative to the null hypotheses, s the standard deviation of the differences on X, and d is the net difference between the null and X. Once the sample distribution of V is established, one easily can obtain a Confidence Interval for d.
It seems that a lot of people know is that d doesn´t exist, I mean.

Luis A. Afonso



Point your RSS reader here for a feed of the latest messages in this topic.

[Privacy Policy] [Terms of Use]

© Drexel University 1994-2014. All Rights Reserved.
The Math Forum is a research and educational enterprise of the Drexel University School of Education.