Rich Ulrich <[EMAIL PROTECTED]> wrote:
> On 21 Jul 2000 07:57:42 GMT, Ron Bloom <[EMAIL PROTECTED]>
> wrote:
> < snip, including citation ... >
>> I was responding to Mr. Ullrich's implication that while,
>> on the one hand the observed "chi-squared" value can be interpreted as
>> a test statistic, on the other hand, the observed "odds ratio" is
>> not a test statistic but a measure of effect size.
>>
>> I was simply pointing out that any combination of a,b,c,d
>> can be, in principle, be used as a "test statistic", since,
> < snip, rest >
> Hey, loosen up....
> Are you still hoping to learn something?
>
Of course I am trying to learn something. That is why I asked
the question in the first place. I was confused as to
whether certain distinctions were "conventions" or were
somehow reflective of a theoretical point that I had missed.
I don't think I "got it" until I had composed my
reply to your first comment.
> I was pointing to an old distinction (I did not invent it), which
> various people have used, and one that answers your question:
> Statistics that incorporate the N are sometimes "test statistics"
> (which I did put in quotes, before), as contrasted to "effect size"
> measurements. They earn those labels because they most readily tell
> you one sort of thing of the other.
Now I think I'm on the same sheet of music.
> And they each tell you something that is a little different, and that
> "scales" differently, from just having the p-level.
This hits the nail on the head
Thanks,
R.b.
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================