Using OMS and Bootstrapping for Logistic Regression

[Michael Kruger]

Dear Members of the List,

   I realize that this is somewhat of an esoteric question, but what
have I got to lose by asking. I have found that one can use the OMS
(Output Management System)
feature in SPSS to create a table of bootstrapped parameter estimates of
the various resamplings. I have been able to do it with multiple linear
regression and save my beta coefficients into a new data file, but am
having some trouble getting the beta coefficients from a logistic
regression to output to a tabled file. With multiple regression
specifiying  /IF COMMANDS ['Regression]) SUBTYPES=['Coefficients'] as an
OMS subcommand does the trick. But I can't figure out what the SUBTYPE [
]  entry should be for COMMANDS ['LOGISTIC REGRESSION']. Anyone who has
any idea, I would appreciate their feedback.

Michael Kruger
"A True Prince"
Statistical Analyst
C.S. Mott Center
Dept. of OB/GYN
Wayne State University School of Medicine
(313)-577-1794

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

[Peck, Jon]

All of the OMS command and subtypes are available via Utilities-OMS identifiers.

You can also see them in the OMS Comtrol Panel on the same menu or by right-clicking in the Viewer outline on an instance of the table.

HTH,
Jon Peck

-----Original Message-----
From: SPSSX(r) Discussion <[hidden email]>
To: [hidden email] <[hidden email]>
Sent: Mon Jul 14 14:33:55 2008
Subject:      [SPSSX-L] Using OMS and Bootstrapping for Logistic Regression

Dear Members of the List,

   I realize that this is somewhat of an esoteric question, but what
have I got to lose by asking. I have found that one can use the OMS
(Output Management System)
feature in SPSS to create a table of bootstrapped parameter estimates of
the various resamplings. I have been able to do it with multiple linear
regression and save my beta coefficients into a new data file, but am
having some trouble getting the beta coefficients from a logistic
regression to output to a tabled file. With multiple regression
specifiying  /IF COMMANDS ['Regression]) SUBTYPES=['Coefficients'] as an
OMS subcommand does the trick. But I can't figure out what the SUBTYPE [
]  entry should be for COMMANDS ['LOGISTIC REGRESSION']. Anyone who has
any idea, I would appreciate their feedback.

Michael Kruger
"A True Prince"
Statistical Analyst
C.S. Mott Center
Dept. of OB/GYN
Wayne State University School of Medicine
(313)-577-1794

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

[Nathaniel Lewis-20]

You could also try using the aggregate command to save out your beta
estimates to file, we have had success for jackknifing applications.

If you are doing > 100 iterations, make sure you do only run 100 at a
time, as SPSS tends to fall over when you push it that hard.

Cheers,

Nathaniel

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Michael Kruger
Sent: Tuesday, 15 July 2008 5:34 AM
To: [hidden email]
Subject: Using OMS and Bootstrapping for Logistic Regression


Dear Members of the List,

   I realize that this is somewhat of an esoteric question, but what
have I got to lose by asking. I have found that one can use the OMS
(Output Management System) feature in SPSS to create a table of
bootstrapped parameter estimates of the various resamplings. I have been
able to do it with multiple linear regression and save my beta
coefficients into a new data file, but am having some trouble getting
the beta coefficients from a logistic regression to output to a tabled
file. With multiple regression specifiying  /IF COMMANDS ['Regression])
SUBTYPES=['Coefficients'] as an OMS subcommand does the trick. But I
can't figure out what the SUBTYPE [ ]  entry should be for COMMANDS
['LOGISTIC REGRESSION']. Anyone who has any idea, I would appreciate
their feedback.

Michael Kruger
"A True Prince"
Statistical Analyst
C.S. Mott Center
Dept. of OB/GYN
Wayne State University School of Medicine
(313)-577-1794

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