No. You must run a new equation without the omitted variable. The b
coefficients you have now are computed controlling for the other variables,
and thus are not the same as the new coefficients you would have when one of
the other variables is omitted.
Hector
-----Original Message-----
From: SPSSX(r) Discussion [mailto:
[hidden email]] On Behalf Of
Piotr Kolodziejczyk
Sent: 16 July 2009 10:09
To:
[hidden email]
Subject: Logistic regression issue
Dear all,
I got equation with 'B' paramteres in logistic regression with Y as
depended and three variables (gender and class) as covariates. Output
is like:
gender:
gender(f): -,531
age:
age(18-30): 0,4
age(31-50): -0,59
age(51+): 1,25
class:
class(A): -,331
class(B) -1,248
class(C) -1,384
Constant: 0,721
So if I want to calculate my (Y) variable depended on each class,
gender, age group I got value of Y.
But I want to get value of Y when one of the variable is ommited.
Could I use any 'mean' of whole factor like class, or age to estimate
value of Y without making new equation?
I'd like to know estmated Y value regardless setting of reference
category for i.e. age.
I'll be grateful for any sugesstions.
Regards,
Piotr
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
