From: Katharine Thayer (firstname.lastname@example.org)
Date: 11/10/00-02:31:11 AM Z
I don't disagree with anything you say here. My point was that if Wayde
wanted to set up a scientific experiment that tested all the variables
against all the other variables, as he's recommending that folks do, he
would be disappointed if he thought one of these correlational models
would yield repeatable or even necessarily interpretable results, in
terms of what quantitative change in which variable or combination of
variables yielded what quantitative change in the outcome. I was hoping
he would work through the whole example and discover this for himself,
as he tried to replicate the results on a second set of data. As I said
in the first post I wrote on this topic, it's my opinion that there
isn't a statistical model that makes sense for our situation and I'm
perfectly comfortable having everyone run tests for their own purposes,
as long as they understand that their results cannot necessarily be
generalized to anyone else.
Judy Seigel wrote:
> On Wed, 8 Nov 2000, Katharine Thayer wrote:
What I've been trying to say over and over again is that
> > your idea that repeatability can be achieved by the methods described in
> > the references you gave is, in my experience, a misplaced faith.
> And perhaps not necessary.... But how you define "repeatability"?
> Is "repeatability" every time without exception, all results cancelled if
> anomaly occurs? And HOW exact is "repeatable" anyway ? Pretty close?
> Similar? Or weigh it on the fingerprint scale (in which case I'm out of
> the game)? My timer is a gralab, not rocket science level. My water is
> available temp from the faucet. My breath may change hourly. BUT I haven't
> found these general conditions rule out other problem solving... which
> suggests another proposition: I'll try to figure this out with my usual
> keen insight, dedicated artistry and general kvetching, under prevailing
> conditions of restrained chaos (which I haven't yet managed to clean up --
> that's first step for the clean room).
> First I'll try to get the discoloration again, because most of my tests
> *don't* have it, tho that may take 6 months -- minimum.
This archive was generated by hypermail 2b30 : 12/01/00-11:46:56 AM Z CST