Plotting Cox Hazard/Survival Functions

[Dottie-2]

I ran a Cox Regression which showed that receiving the "treatment" (dummy
vbl indicating participation in a prevention program)reduces the hazard
for committing a school discipline offense (B = -.305; Exp(B)= .737;
p<.000)).

However, when I plot the survival function for treatment vs. controls (y-
axis = increasing survival, x-axis = time) the treatment group is shown
the lower of the two lines (i.e., indicating lower survival)?

Similarly, when I plot hazard, the treatment group is the higher of the
two lines (i.e., indicating greater hazard).

What am I missing here?

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@... (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

[Reutter, Alex]

How was the dummy variable coded?  If, for example, 0=did not participate and 1=did participate, then Coxreg has likely coded the indicators so that your coefficient of -.305 corresponds to membership in the "did not participate" group.  Look at the Categorical Variable Codings table and compare to the Variables in the Equation table to see exactly what happened.

Cheers,
Alex


-----Original Message-----
From: SPSSX(r) Discussion [mailto:SPSSX-L@...] On Behalf Of Dottie
Sent: Wednesday, July 02, 2008 2:26 PM
To: SPSSX-L@...
Subject: Plotting Cox Hazard/Survival Functions

I ran a Cox Regression which showed that receiving the "treatment" (dummy
vbl indicating participation in a prevention program)reduces the hazard
for committing a school discipline offense (B = -.305; Exp(B)= .737;
p<.000)).

However, when I plot the survival function for treatment vs. controls (y-
axis = increasing survival, x-axis = time) the treatment group is shown
the lower of the two lines (i.e., indicating lower survival)?

Similarly, when I plot hazard, the treatment group is the higher of the
two lines (i.e., indicating greater hazard).

What am I missing here?

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@... (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
LISTSERV@... (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