Re: Regression significance conundrum


Andy Spragg wrote:
> On 27 Oct 2005 16:55:26 -0700, "Reef Fish"
> <Large_Nassau_Grouper@xxxxxxxxx> wrote:
>
> >Andy Spragg wrote:
>
> >> On 22 Oct 2005 13:46:29 -0700, "Reef Fish"
> >> <Large_Nassau_Grouper@xxxxxxxxx> wrote:
>
> >> >The use of the term "correlation" rather than "linear functional
> >> >relation" seems to be still a remnant of your old ills.
> >>
> >> No, I don't think so. I'm very happy with the term "correlation"
> >> (although I should have said "linear correlation"). A "function" is an
> >> unambiguous specification of an output given an input. If we're going
> >> to be picky about terminology, I suppose I would say that if y is
> >> linearly correlated with x, then there is a linear functional
> >> relationship between yhat and x. So to speak.
>
> >No, you're TOTALLY wrong.
>
> (snip)
>
> You're right; whoops. Let me try again:
>
> If we're going to be picky about terminology, I suppose I would say
> that if y is linearly correlated with x, and we use that correlation
> as the basis for fitting a linear model, then there is a linear
> functional relationship between yhat and x. So to speak.

You're still TOTALLY wrong, and you've exposed one more of your WRONGS.

The fitting of a "linear model" has NOTHING to do with the linearity
of the functional relationship between X and Y. See

http://tinyurl.com/bh9ft

"What are LINEAR or LINEAR REGRESSION models?" is a subthread in
sci.stat.math that ran over 100 posts; there are several OTHER
threads in sci.stat.math dealing with that same subject.

You apparently missed it ALL, confusing a "linear regression model"
with a "linear function model between Y and X".


>
> The point I was trying to make was that I was not wrong to use the
> term "correlation" instead of "linear functional relationship". If two
> variables are correlated, they are not "linearly functionally related"
> or indeed "anything else functionally related". They are statistically
> related. The functional relationship is in the model that is fitted.
> (IMO, YMMV etc).

You were WRONG, for more reasons than one. You have now exposed
your other WRONG, about not knowing what a linear regression model is!

-- Bob.

At sunny Waikiki beach, Honolulu.



>
> >> Thank you for your thoughts. I think previous respondents had already,
> >> more clearly and without animosity, pointed out the fresh errors of my
> >> ways.
>
> >I gave it to you straight, about where you erred, just I told you
> >NOW about how you SERIOUSLY erred about what correlation measures.
> >
> >You OBVIOUSLY hadn't learned what I told you now, from any of your
> >previous respondents, have you?
>
> No, because what we are talking about now is the one little bit that
> no-one else did say, namely this terminological business about whether
> we should say "linear functional relationship" instead of
> "correlation". I think that's because no-one else has a problem with
> just using the term "correlation", BICBW.
>
> I thought, and think, that you are wrong. But you were correct in
> picking me up for missing out one clause when I attempted to point out
> why.

.

Relevant Pages

  • Re: Simple Sagnac
    ... In the frame of the train, ... > |> There is no linear relationship between t and tau and you've ... > |> cuckoo transforms are linear? ... > | A linear function of two variables is the sum of a constant times ...
    (sci.physics.relativity)
  • Re: how to find a1,a2...an for the goal value?
    ... set of parameters there is a linear function of x for which the mean ... square difference between it and gis a minimum. ... between gand a best-fitting linear function in the above sense. ...
    (comp.soft-sys.matlab)
  • Re: Why momentum is a cotangent vector?
    ... elements of fibers are called momentum vectors. ... I have no special knowledge here, but in general, co-vectors are linear ... L is a function from velocities to reals. ... To get a linear function out of it, ...
    (sci.physics)
  • Re: FIGHT! FIGHT! FIGHT!
    ... requirements of a linear function when applied to network theory." ... Linear means the graph of the function is a straight line. ... In y = mx + b, m is a constant determining the slope of the line. ...
    (rec.radio.amateur.antenna)
  • Re: Richard Ulrich is being a pest AGAIN! (was Re: r-Squared Question)
    ... I suppose it would be just-as-fair if I described Bob Ling ... that "Richard Ulrich does not even know what a LINEAR ... called an "abecedarian linear regression model". ... What is the cost, or the loss, of doing so? ...
    (sci.stat.math)