Date: Dec 7, 2012 1:17 AM
Author: Ray Koopman
Subject: Re: So do we need to "Bonferroni-correct" in this case

On Dec 6, 7:14 am, djh <halitsk...@att.net> wrote:
> According to your estimates (thank you very much for doing them!),
> below are our potentially acceptable p's. Do we need to Bonferroni-
> correct before proceeding further based on these p's? I would argue
> no ... but of course it's your call ...
>
> Het
> N1 Aube H H L L H H H H L L L L 0
> Aubqe H H L L H H H H L L L L 0
>
> N3 Aube L L H H L L H H H H L L 0
> Aubqe L L H H L L H H H H L L 0
>
> p ~= .021822 for het 0
>
> L-H
> Het
> N1 Auq L H L H L H L H L H L H 6
> Aubu L H L H H H L H L H H H 4
>
> N2 Auq L H H H L H H H L H L H 4
> Aubu H H L H L H H H L H L H 4
>
> N3 Auq L H H H H H L H L H L H 4
> Aubu L H H H L L L H L H L H 4
> Aubqu L H H H H H L H L H L H 4
>
> p ~= .001077 for L-H het 6
> p ~= .048634 for L-H het 4
>
> H-L
> Het
> N1 AubC H L H L H L H L H L H L 6
> AubqC H L H L H L H L H L H L 6
>
> N2 AubC H L H L H L H L H L H L 6
> AubqC H L H L H L H L H L H L 6
>
> p ~= .001077 for H-L het 6


1. Het = 3 (posted 12/5 @ 12:00) is impossible. Check your data.

2. Why do you present results from the regressions of c on both
(u,e,u*e) and (u,e,u*e,u^2) for the same data? It is unusual to
consider the results of both analyses, except for the purpose of
deciding which model to use. Have you checked the significance of
the quadratic term? Does its inclusion reduce the Standard Error of
Prediction (SEP) substantially? (Those are two conceptually separate
issues.)

3. What does "Het" *mean*? I'm always suspicious of anything that
starts with arbitrary dichotomization. What are you trying to show?