Re: Simple Q. about co-efficient in coded regression equation

On Sun, 2 Nov 2008 06:05:20 -0800 (PST), "geetha.shree@xxxxxxxxx"
<geetha.shree@xxxxxxxxx> wrote:

Hi Mr. Ulrich,
Thanks for your reply.
What I meant by Coded regression was , I did "Centering of Predictor"
variable and continued the regression analysis for the new predictor
variables which are centered. And the resulting equation would be
interpreted the same way as in the original regression equation
rght ?

I think you need to work on precision of writing.

I don't see how Coded regression with some reference to
-1 and 1 becomes Centering of predictor. However, if you
only subract a constant, the regression coefficients are
not changed, except for the Constant.


b) Just want to confirm ,are you suggesting me tht Most of the major
Software use "Stepwise Regression by default in Regression Analysis ?
[snip, rest]

And work on reading. I said, in shorter terms -- I don't know
and don't much care, because "stepwise is stupid". Usually.

I said, if you want to know, try it, or read some documentation.

--

Rich Ulrich


Thanks,

On Nov 1, 6:00 pm, RichUlrich <rich.ulr...@xxxxxxxxxxx> wrote:
On Sat, 1 Nov 2008 13:59:34 -0700 (PDT), "geetha.sh...@xxxxxxxxx"



<geetha.sh...@xxxxxxxxx> wrote:
Hi
I have two equation :
a) From Regression Analysis
b) Coded Regression i.e., after coding the predictor variables ( 1 to
-1 )

Case 1:
a) Y =  11.4 - 2.07 X1 - 2.68 X2 - 0.624 X3 - 85.3 X4 - 76.2 X5
b) Y = 7.45 - 2.07 X1 - 2.68 X2 - 0.624 X3 - 85.3 X4 - 76.2 X5

Case 2:
a) Y = - 5.50 + 0.723 X1 + 3.09 X2 - 3.74 X3 + 5.21 X4 + 70.2 X5
b) Y = - 4.75 + 0.591 X1 + 0.78 X2 - 3.58 X3 + 4.11 X4 + 91.5 X5

Q1 : I just want to know  if we consider Coded regression equation in
either of the two cases, do we  consider  the highest co-efficient as
the variable which has more effect on the model  or vice verse ?

"Coded regression" is not a conventional term for me, so I
don't know whether you are referring to dummy variables with
two values, or if you are constricting the ranges.

In either case, the meaning of the coefficient is the regressed
effect of a 1-point change in the predictor.  In either case,
the *variance* of the predictor has an influence on which
predictor has the larger point-impact on the *varniance* of
the prediction, across all cases.  The standardized beta given
by most programs will show the largest.   The sum of (beta*r_0)
where r_0  is the zero-order correlation creates a "partition"
that adds up the  total R^2, when all the terms are positive.
It is not used much because, too often, there are "suppressors"
with negative contributions to the sum.



Q 2 : By DEFAULT, what method does Minitab, Spss software use for
regression analysis if we dont specify ( is it backward, forward,
stepwise ).

"Stepwise" in any direction is such a dubious idea that you
should only engage in it if you have specific knowledge and
intentions.  If you are really interested, look at documentation
or try it.    Google groups <group:sci.stat.*  stepwise author:ulrich>
for threads on the subject.

--
Rich Ulrich
.

Relevant Pages

  • Re: Regression with Dichotomous Dependent Variables
    ... What immediately occurs to me is "discriminant function" for two ... and that may be why logistic regression is becoming ... I don't know how it works out for logistic regression, ... create a multi-variable predictor *composite* score. ...
    (sci.stat.math)
  • Re: Simple Q. about co-efficient in coded regression equation
    ... What I meant by Coded regression was, I did "Centering of Predictor" ... Software use "Stepwise Regression by default in Regression Analysis? ...
    (sci.stat.math)
  • Re: Regression with Dichotomous Dependent Variables
    ... A discriminant function will predict which of two groups an observation likely belongs to ... Logistic regression makes no assumptions about the predictors. ... the linear regression approach violates a cardinal assumption -- the distribution of the disturbances is now strongly related to the predictor values. ...
    (sci.stat.math)
  • Re: Simple Question on Forward and backward regression
    ... through BACKWARD Regression must be reflected in FORWARD Regression as ... I found the most significant Predictor variables ... of stepwise regression, let alone forward/backward stepwise, which I know is flammable in some quarters. ... The F test in the forward direction to decide whether to add X1 to X3+X5 should be the same as the F test in the backward direction to decide whether to turf X1 out of X1+X3+X5. ...
    (sci.stat.math)
  • sample size for logistic regression (Babyak/Peduzzi papers)
    ... What I have understood is that in logistic regression I need at least 10 observation per predictor variable and the sample size that must be considered is the smallest number between the events and the non-events, for example if I have 1000 observations but only 50 events my sample size is 50 so I "can" use about 5 predictors. ... - a categorical predictor that becomes for example 5 dummy variables counts as ONE or FIVE predictors? ... Level 1: 10 EVENTS 20 NONEVENTS ...
    (sci.stat.math)
Loading