Multiple imputation question

[Joost van Ginkel]

Dear Kirill,

 

I can give you the references of my most relevant contributions to the field:

 

Van Wingerde, B. & Van Ginkel, J. (2021). SPSS syntax for combining results of Principal Component Analysis of multiply imputed data sets using Generalized Procrustes Analysis. Applied Psychological Measurement. doi: 10.1177/0146621621990757

 

Van Ginkel, J.R. (2020). Standardized regression coefficients and newly proposed estimators for R² in multiply imputed data. Psychometrika, 85, 185–205. doi: 10.1007/s11336-020-09696-4

 

Van Ginkel, J.R. (2019). Significance tests and estimates for R² for multiple regression in multiply imputed datasets: a cautionary note on earlier findings, and alternative solutions. Multivariate Behavioral Research, 54, 514-529. doi: 10.1080/00273171.2018.1540967

 

Van Ginkel, J.R. Linting, M., Rippe, R.C.A., & Van der Voort, A. (2020). Rebutting existing misconceptions about multiple imputation as a method for handling missing data. Journal of Personality Assessment, 102, 297-308. doi: 10.1080/00223891.2018.1530680

 

Van Ginkel, J.R. & Kroonenberg, P.M. (2014a). Using Generalized Procrustes analysis for multiple imputation in principal component analysis. Journal of Classification, 31, 242-269. doi: 10.1007/s00357-014-9154-y

 

Van Ginkel, J.R. & Kroonenberg, P.M. (2014b). Analysis of variance of multiply imputed data. Multivariate Behavioral Research, 49, 78-91. doi:10.1080/00273171.2013.855890

 

Please, let me know if there are any references that you can’t get access to you. I could send them to you personally.

 

Best,

 

Joost

 

From: Kirill Orlov <[hidden email]>
Sent: Wednesday, April 7, 2021 2:11 PM
To: Ginkel, J.R. van <[hidden email]>; [hidden email]
Subject: Re: Multiple imputation question

 

Oh, bravo!
Are your paper and other papers mentioned here available for free somewhere? Do you have a site where one can download them?

07.04.2021 15:03, Ginkel, J.R. van пишет:

In case of Exploratory Factor Analysis without any statistical tests you are right. As a matter of fact, Van Ginkel and Kroonenberg (2014) studied this option (averaging a correlation matrix) in the context of pooling the results of PCA, together with two other methods (averaging component loadings and Generalized Procrustes Analysis).

 

From: SPSSX(r) Discussion [hidden email] On Behalf Of Kirill Orlov
Sent: Wednesday, April 7, 2021 1:58 PM
To: [hidden email]
Subject: Re: Multiple imputation question

 

 

 

===================== 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]

I have frequently used factor analysis for summative scales in the years
before MI was made available.

A very early lesson was to tighten administration to prevent missing data.  
Another early lesson was that items that had missing data often had
substantive problems such as respondents skipping questions that were too
private or that were confusing.

One potential problem with using items for MI is that it may be harder to
get cleanly loading items to maximize divergent validity.

There are reasons why scale development conventionally starts with a lot
more candidate items than would be administered than one wishes to have
retained items.



-----
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

[Jeff A]

Kirill, I was just about to ask you if you had a particular macro you would recommend, but after a google search I see that you have one. ...just downloaded it and your description of the procedure in the accompanying word document looks relatively simple to follow (although I don't think I would want to try to understand what that macro syntax is going exactly).

It's not directly related to my original question, but I'll definitely have a look at the macro and hot-deck more specifically since I've never actually used it myself. It appears to be a much easier way of handing missing data than MI and can be used easily with any type of analysis. Given that the type of data I generally use is subject to a great deal of measurement error, I'm assuming that HD will be more than adequate for the majority of my uses and that MI might be unnecessary overkill.

Thanks

Jeff




-----Original Message-----
From: SPSSX(r) Discussion <[hidden email]> On Behalf Of Kirill Orlov
Sent: Wednesday, April 7, 2021 10:08 PM
To: [hidden email]
Subject: Re: Multiple imputation question

Jeff, although this thread is not discussing hot-deck imputation, I want to remind to you that option too.
Hot-deck method is not unpopular in social sciences, especially in surveys.
SPSS does not have in-build command for hot-deck, but there exist macros.

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

[Kirill Orlov]

Jeff, If you are using SPSS 25 or later version, you can also install
the KO_macros.spe extension which allows to run some of my macros,
including the HD imputation macros (there are two), from menu dialogs.
My web-page is here: http://www.spsstools.net/en/KO-spssmacros.
Thanks for the interest.


08.04.2021 2:51, Jeff A пишет:
> Kirill, I was just about to ask you if you had a particular macro you would recommend, but after a google search I see that you have one. ...just downloaded it and your description of the procedure in the accompanying word document looks relatively simple to follow (although I don't think I would want to try to understand what that macro syntax is going exactly).
>
> It's not directly related to my original question, but I'll definitely have a look at the macro and hot-deck more specifically since I've never actually used it myself. It appears to be a much easier way of handing missing data than MI and can be used easily with any type of analysis. Given that the type of data I generally use is subject to a great deal of measurement error, I'm assuming that HD will be more than adequate for the majority of my uses and that MI might be unnecessary overkill.
>
> Thanks
>

=====================
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]

<begin soapbox>
When using heuristic methods, one should use several methods and see whether
the choice of the method makes a substantive difference in the reasoning the
data are being used to support.
<end soapbox>



-----
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