Missing Value Analysis Question

[Salbod]

Hi Friends,

This is the first time I’m using MVA and I’m not sure if the output of a regression analysis was complete. I implemented the Missing Values Analysis procedure with my dataset: Missing Value Analysis->Multiple Imputation->Impute Missing Data Values (Accepted default=5). I ran a Regression analysis (Linear) and expected Pooled Anova and Model Summary results. I got results for the original dataset plus the 5 imputations, but no pooled analyses.  

Any explanation will be appreciated.

TIA, Stephen Salbod, Pace University, NYC

[Joost van Ginkel]

Dear Stephen,

See the reference below:

Van Ginkel, J.R. & Kroonenberg, P.M. (2014). Analysis of variance of multiply imputed data. Multivariate Behavioral Research, 49, 78-91.

The paper also describes how to do it in SPSS using a macro that was programmed by me. The link to this macro may be found in the paper. If you have any further questions about the procedure, please let me know.

Best regards,

Joost van Ginkel

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Salbod
Sent: woensdag 25 maart 2015 15:12
To: [hidden email]
Subject: Missing Value Analysis Question

Hi Friends,

This is the first time I’m using MVA and I’m not sure if the output of a regression analysis was complete. I implemented the Missing Values Analysis procedure with my dataset: Missing Value Analysis->Multiple
Imputation->Impute Missing Data Values (Accepted default=5). I ran a
Regression analysis (Linear) and expected Pooled Anova and Model Summary results. I got results for the original dataset plus the 5 imputations, but no pooled analyses.  

Any explanation will be appreciated.

TIA, Stephen Salbod, Pace University, NYC



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Missing-Value-Analysis-Question-tp5729052.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

[cbrawner]

First, this group is a great resource. The best way to get assistance is to provide the syntax that generated your output.

That aside, in my limited experience using imputed datasets in SPSSv21, I was not able to get pooled analyses when I was implementing a stepwise variable selection procedure in Cox regression. I was able to get pooled analyses when all variables were entered into the model.

Best wishes,
Clinton

[Joost van Ginkel]

Dear Clinton,

The stepwise procedure doesn't work because for some imputed data sets other predictors may have been selected than for others, thus resulting in different final models. Actually SPSS should select variables at the pooled level (based on the pooled significance tests), rather than run a stepwise procedure per imputed data set. Maybe it is possible to carry out the stepwise procedure manually (e.g., add a predictor manually using Enter, see if its pooled test is significant, and then add a second one and see if that one is significant, and so on).

Best, Joost

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of cbrawner
Sent: woensdag 25 maart 2015 15:22
To: [hidden email]
Subject: Re: Missing Value Analysis Question

First, this group is a great resource. The best way to get assistance is to provide the syntax that generated your output.

That aside, in my limited experience using imputed datasets in SPSSv21, I was not able to get pooled analyses when I was implementing a stepwise variable selection procedure in Cox regression. I was able to get pooled analyses when all variables were entered into the model.

Best wishes,
Clinton



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Missing-Value-Analysis-Question-tp5729052p5729055.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