Decision tree

[Moshe Marko]

Hi,

I am trying to use the classification tree analysis to test which one of
five outcome measures can best predict persons with physical disability
(dependent variable which is a categorical variable [where 1 = yes
disability, 0 = no disability] was categorized. When I run the program it
tells me that none of the independent outcome measures can be included in
the model. That is despite the fact that the independent outcome measures
highly correlate with the dependent variable. I looked at the tutorial and
could not figure out what was I doing wrong

Thanks for the help

Moshe

Moshe Marko, PT, DPT, MHS, OCS, CSCS
Assistant Professor
Department of Physical Therapy Education
College of Health Professions
SUNY Upstate Medical University
Room 2232  Silverman Hall
750 Adams Street
Syracuse, NY 13210-1834
315 464 6577
FAX [phone]
[hidden email]

=====================
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

[John McConnell-2]

Hi Moshe

If i understand you correctly it sounds like you have one
"Target"(/Dependent) variable which is Physical Disability?

And one or more "outcome" measures as "inputs" (/independents)?

My best guess is that this is a sample size issue. Bu default - and this
is set in the Criteria button on the Decision Tree Dialog - you have to
have 100 cases in the parent node and 50 in the child.

I understand that confidentiality may not permit this but if you can
send the data over i can probably help more

And if you don't have too many vars a good old crosstab with a chi
square might give you the info you need. And it you have the Tables
module you will get a Column Proportion Test that is more specific than
the Chi Square.

Hope this helps

John

Analytical People


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Moshe Marko
Sent: 09 July 2010 16:14
To: [hidden email]
Subject: Decision tree

Hi,

I am trying to use the classification tree analysis to test which one of
five outcome measures can best predict persons with physical disability
(dependent variable which is a categorical variable [where 1 = yes
disability, 0 = no disability] was categorized. When I run the program
it
tells me that none of the independent outcome measures can be included
in
the model. That is despite the fact that the independent outcome
measures
highly correlate with the dependent variable. I looked at the tutorial
and
could not figure out what was I doing wrong

Thanks for the help

Moshe

Moshe Marko, PT, DPT, MHS, OCS, CSCS
Assistant Professor
Department of Physical Therapy Education
College of Health Professions
SUNY Upstate Medical University
Room 2232  Silverman Hall
750 Adams Street
Syracuse, NY 13210-1834
315 464 6577
FAX [phone]
[hidden email]

=====================
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