[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Evidence column in -mlogtest-

From   Maarten buis <[email protected]>
To   [email protected]
Subject   Re: st: Evidence column in -mlogtest-
Date   Tue, 18 Sep 2007 15:47:45 +0100 (BST)

--- Reza Daniels <[email protected]> wrote:
> The column is labelled "Evidence", and in both rows it states that
> there is evidence for the null that IIA is satisfied. However, the
> p-values suggest that only one alternative supports the null. 

> **** Small-Hsiao tests of IIA assumption
>  Ho: Odds(Outcome-J vs Outcome-K) are independent of other
> alternatives.
>  Omitted |  lnL(full)  lnL(omit)    chi2   df   P>chi2   evidence
> ---------+---------------------------------------------------------
>        2 |   -973.735   -939.394  68.683   52    0.060   for Ho    
>        3 |  -2506.763  -2475.979  61.567   52    0.171   for Ho    
> -------------------------------------------------------------------

In both rows the p-value is more than .05, so in both case the Null
hypothesis is not rejected. I think the terminology in this case is
poorly chosen: with frequentist testing the the more appropriate
terminology is that we either reject or not reject the null hypothesis.
In a Bayesian framework we get evidence for a hypothesis. In Stata you
will rarely encounter this perspective, the only exceptions are when
dealing with BICs or when you are using the winbugsfromstata package by
J. Thompson, T. Palmer, and S. Moreno. 

Hope this helps,

Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

Want ideas for reducing your carbon footprint? Visit Yahoo! For Good
*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index