Why multiple models are made in linear regression?

[ssuave]

Hi there,

 I wonder how multiple models are made in a linear regression model? See for example in the attached file, 4 models were made. I wonder why not 2 or why not 5?
How?

Linear_regression_svm.pdf

[parisec]

This is because you specified the model to be backwards regression.
Model 1: All of the variables are entered;
Model 2: the variable with the highest F-value (=>100) was pulled out;
Model 3: with VAR17 out, the remaining variables are examined and the one with highest F-value was pulled out;
Model 4: same as model 3.

The process stopped because after pulling all of these variables out, there were no more variables with an F value =>100.

Carol

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of ssuave
Sent: Thursday, October 18, 2012 7:35 AM
To: [hidden email]
Subject: Why multiple models are made in linear regression?

Hi there,

 I wonder how multiple models are made in a linear regression model? See for example in the attached file, 4 models were made. I wonder why not 2 or why not 5?
How?

Linear_regression_svm.pdf
<http://spssx-discussion.1045642.n5.nabble.com/file/n5715719/Linear_regression_svm.pdf>



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Why-multiple-models-are-made-in-linear-regression-tp5715719.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

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

[Bruce Weaver]

And before Rich U does it, let me direct the OP to this list of problems with algorithmic variable selection methods:

   http://www.stata.com/support/faqs/statistics/stepwise-regression-problems/

;-)

parisec wrote

This is because you specified the model to be backwards regression.
Model 1: All of the variables are entered;
Model 2: the variable with the highest F-value (=>100) was pulled out;
Model 3: with VAR17 out, the remaining variables are examined and the one with highest F-value was pulled out;
Model 4: same as model 3.

The process stopped because after pulling all of these variables out, there were no more variables with an F value =>100.

Carol

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of ssuave
Sent: Thursday, October 18, 2012 7:35 AM
To: [hidden email]
Subject: Why multiple models are made in linear regression?

Hi there,

 I wonder how multiple models are made in a linear regression model? See for example in the attached file, 4 models were made. I wonder why not 2 or why not 5?
How?

Linear_regression_svm.pdf
<http://spssx-discussion.1045642.n5.nabble.com/file/n5715719/Linear_regression_svm.pdf>



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Why-multiple-models-are-made-in-linear-regression-tp5715719.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

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