Does anybody now what to do with a warning when performing multiple imputation in SPSS?

classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|

Does anybody now what to do with a warning when performing multiple imputation in SPSS?

shirisha
Hello,

I'm trying to run multiple imputation on SPSS 25 with defined constraints,
and am getting the following warning that stops the execution of the
process:

"After 100 draws, the imputation algorithm cannot find an imputed value
under the constraints for variable XXX. Check the minimum and maximum values
specified to determine that they are reasonable, and consider raising the
number of draws allowed."

I've checked the min and max values and raised the number of draws, but am
still getting the same error with the same or different variable.

The thing is that if I don't define constraints, let SPSS "figure out" min
and max itself, SPSS imputes out of the range values. Let's say if min and
max for the variable is 1 and 5, imputations can be 0.4 or 6.2.
Thank you!



--
Sent from: http://spssx-discussion.1045642.n5.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
Reply | Threaded
Open this post in threaded view
|

Re: Does anybody now what to do with a warning when performing multiple imputation in SPSS?

Ki Park
I would recommend posting the code and more in-depth description of your data
and variable(s). If you are using the dropdown menu for this imputation, you
could use "paste" instead of "run" and get the syntax/code that you are
specifying.
Also, paste all the text in the warning, so it may give some clue of what is
happening.
Thanks.




--
Sent from: http://spssx-discussion.1045642.n5.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
Reply | Threaded
Open this post in threaded view
|

Re: Does anybody now what to do with a warning when performing multiple imputation in SPSS?

Rich Ulrich
In reply to this post by shirisha
My first guess is that you are dealing with variables that need
transformations in order to be "equal interval measures" of
whatever you are measuring.  That's the easiest explanation.
Do you have variables with extreme skew or outliers?

If you are only concerned that the imputations are out of range,
well, that's the theoretical shortcoming of discriminant function
compared to logistic regression - Predictions can be out of range.
Pragmatically, here, I would think you might compute them and
recode them to the range.

But I don't know at all what other folks consider acceptable.

--
Rich Ulrich


From: SPSSX(r) Discussion <[hidden email]> on behalf of shirisha <[hidden email]>
Sent: Thursday, February 7, 2019 9:06 PM
To: [hidden email]
Subject: Does anybody now what to do with a warning when performing multiple imputation in SPSS?
 
Hello,

I'm trying to run multiple imputation on SPSS 25 with defined constraints,
and am getting the following warning that stops the execution of the
process:

"After 100 draws, the imputation algorithm cannot find an imputed value
under the constraints for variable XXX. Check the minimum and maximum values
specified to determine that they are reasonable, and consider raising the
number of draws allowed."

I've checked the min and max values and raised the number of draws, but am
still getting the same error with the same or different variable.

The thing is that if I don't define constraints, let SPSS "figure out" min
and max itself, SPSS imputes out of the range values. Let's say if min and
max for the variable is 1 and 5, imputations can be 0.4 or 6.2.
Thank you!



--
Sent from: http://spssx-discussion.1045642.n5.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
Reply | Threaded
Open this post in threaded view
|

Re: Does anybody now what to do with a warning when performing multiple imputation in SPSS?

shirisha
In reply to this post by Ki Park
MI-c_Output.spv
<http://spssx-discussion.1045642.n5.nabble.com/file/t341571/MI-c_Output.spv>  

Thank you, Ki, for your interest to help. I just uploaded the SPSS output
file with the code and the message I get there. Hope it is helpful.

Thank you,
SK



--
Sent from: http://spssx-discussion.1045642.n5.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
Reply | Threaded
Open this post in threaded view
|

Re: Does anybody now what to do with a warning when performing multiple imputation in SPSS?

shirisha
In reply to this post by Rich Ulrich
Thank you, Rich, for your message.

You are correct, 12 scales out 20 in my study are skewed. So, does that mean
that I need to transform the data before imputing?

Can you share a bit more how I can compute and recode them to the range?

Thank you.
SK



--
Sent from: http://spssx-discussion.1045642.n5.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
Reply | Threaded
Open this post in threaded view
|

Re: Does anybody now what to do with a warning when performing multiple imputation in SPSS?

Bruce Weaver
Administrator
In reply to this post by shirisha
Hello Shirisha.  As odd as it might sound, not all group members have access
to SPSS, and those who do not will not be able to open your .spv output
file.  If you are going to attach output, I suggest exporting it to another
format first (e.g, PDF or Excel).  

Also, members who participate via the UGA mailing list will not receive
attached files.  They'll need to visit this thread in the Nabble archive
(http://spssx-discussion.1045642.n5.nabble.com/Does-anybody-now-what-to-do-with-a-warning-when-performing-multiple-imputation-in-SPSS-td5737384.html)
should they wish to download the file(s).  

Finally, the command syntax is just text, so could be pasted into the body
of your post.  

HTH.



shirisha wrote

> MI-c_Output.spv
> &lt;http://spssx-discussion.1045642.n5.nabble.com/file/t341571/MI-c_Output.spv&gt; 
>
> Thank you, Ki, for your interest to help. I just uploaded the SPSS output
> file with the code and the message I get there. Hope it is helpful.
>
> Thank you,
> SK
>
>
>
> --
> Sent from: http://spssx-discussion.1045642.n5.nabble.com/
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> LISTSERV@.UGA

>  (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
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
Sent from: http://spssx-discussion.1045642.n5.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
--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.
Reply | Threaded
Open this post in threaded view
|

Re: Does anybody now what to do with a warning when performing multiple imputation in SPSS?

Art Kendall
In reply to this post by shirisha
Please explain in more detail what you are trying to do.
What constructs are you measuring?

How are you using the measures you are trying to repair?

Scales?
Do you have summative scales (i.e., using the total or mean of a set of
items)?
The are most commonly used with attitudes, persality measures etc.

Why is the data missing?






-----
Art Kendall
Social Research Consultants
--
Sent from: http://spssx-discussion.1045642.n5.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
Art Kendall
Social Research Consultants
Reply | Threaded
Open this post in threaded view
|

Re: Does anybody now what to do with a warning when performing multiple imputation in SPSS?

Rich Ulrich
In reply to this post by shirisha
Sent: Friday, February 8, 2019 12:52 PM
To: [hidden email]
Subject: Re: Does anybody now what to do with a warning when performing multiple imputation in SPSS?
 
SK
> Thank you, Rich, for your message.

> You are correct, 12 scales out 20 in my study are skewed. So, does that mean
> that I need to transform the data before imputing?

"equal interval" is the important characteristic for data for models and tests.
I always considered whether variables needed transformations - though, at
times, the main investigator ruled that out because there was too much (bad)
conventional practice to justify a  minor gain.  If you don't have equal intervals,
you can't expect homogeneity of variance ... putting the testing into question.
And how do you construct a reasonable linear model when "one unit" is not the
same at both ends of the scoring?

Still, "scales", if these are summative rating scales, are somewhat less prone to
needing transformations that those measures in natural units (which some
people naively take as "equal") like blood titers, distances, times, or counts. 
But with low "counts" in a scale, you might want square root (as with counts).
The bigger treatment of nonlinearity for scales is the IRT approach, which uses
the logistic transformation since the scales are bounded at both ends.

Skewness and outliers or other non-normality are warning signs that you may
not have equal intervals. If you mention the actual variables and their distributions,
readers here would have more to advise from.

> Can you share a bit more how I can compute and recode them to the range?

If you are imputing a single sample, you take the imputed value and recode.
Multiple imputation?  - Sorry, I don't know.  What happens if you expand the
Min/Max range, and just go with it?

I should say again, I'm winging it here, without actual experience and very little
reading in the area.

--
Rich Ulrich

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