Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.
|
|
|
|
Re: F-test for comparing variances?
Posted:
Nov 8, 2006 11:01 PM
|
|
On 8 Nov 2006 11:08:22 -0800, "Reef Fish"
<large_nassua_grouper@yahoo.com> wrote:
>
> Richard Ulrich wrote:
> > On 5 Nov 2006 17:55:46 -0800, "Reef Fish"
> > <large_nassua_grouper@yahoo.com> wrote:
> > >
[snip]
RU > >
> > umm. You still need to get a test value, subtract, then double....
> > which I described usefully. Bob was not paying attention, too eager
> > to criticize.
RF >
> WRONG, WRONG, WRONG.
>
> 1. For testing with a given alpha, such as a two-tailed test with F,
> you find F(.025) and F(.975) before you get the test value,
> because
> you DON'T have to see if the numerator is greater or smaller. You
> KNOW the degrees of freedom of the F.
I can apologize, and clarify.
No doubt I confused Bob by saying "test value" instead
of saying "two-tailed p-value."
If you want a two-tailed p-value -- which is the modern
standard, instead of simply 'flagging' the test result --
you need to double the smaller extreme.
Here are a *couple* of conventions -- presenting p-values;
and doubling a one-tailed value to get a two-tailed value.
[snip]
RU >
> > By now, he should recognize that he *often* does not understand
> > what's being said. Why else is it, that every 'fight' in the stat
> > groups features Bob?
RF >
> Everytime you say I don't understand, it's always because I was at
> several steps AHEAD of you and understand every word of yours to
> be ERRORS resulting from your lack of education.
Your next step, Bob, is to consider how to compute the "p-value",
which everyone wants these days. Catch up.
[snip]
>
> You REMAIN ignorant, and a malpracticing Quack, Richard Ulrich.
That remains absurd. Stream-of-consciousness garbage.
>
> That may be an insult you deserve -- but hardly gratuitous!
Yes. Gratuitous. Nobody asks for this, and 10,000 readers do
not want to see it. *If* I were so obviously 'ignorant', nobody
would *need* Reef Fish Bob Ling to belabor the point.
The only 'end' that it seems to serve is an ongoing intention
to bully and intimidate. Bad net-citizenship.
--
Rich Ulrich, wpilib@pitt.edu
http://www.pitt.edu/~wpilib/index.html
|
|
|
|
