Hello,
I was wondering if anyone could help me with a logistic regression I
am carrying out in SPSS.
My DV is behaviour (smoked in the last 3 months / not smoked in
the last three months)
My IV's are intention, control, and whether or not the participant
was in one of four experimental conditions (II, II-control, SE or SE-
control)
I have created four dummy variables -
COND1) In the II condition or not (0,1)
COND2) In the II-control condition or not (0,1)
COND3) In the SE condition or not (0,1)
and
COND4) In the SE-control condition or not (0,1)
This is my syntax;
LOGI REG VARIABLES= BEHAVIOR WITH INTENTI CONTROL
COND1 COND2 COND3 COND4
/CATEGORICAL= COND1 COND2 COND3 COND4
/METHOD=ENTER INTENTI
/METHOD=ENTER CONTROL
/METHOD=ENTER COND1 COND2 COND3 COND4.
but when I run it, it leaves off one of my variables, giving me this
message;
> Redundancies in Design Matrix:
>
> COND4(1) = 3 - COND1(1) - COND2(1) - COND3(1)
> The variable COND4 will have its degrees of freedom
> reduced from 1 to 0
Is there any way that i can stop it from doing this? I guess it's
because anyone who is not in the other three conditions must be in
th 4th one, but I'd still like to see this one in the output....
Thanks in advance,
Brian
-------------------------------------------------------------
Brian McMillan
Mother and Infant Unit
22 Hyde Terrace
University of Leeds
Leeds
LS2 9LN
tel: 0113 2331575
mob: 0779 0866573
--------------------------------------------------------------
|