SPSS Statistics

 View Only
  • 1.  Iteration history for multiple imputation

    Posted Tue March 16, 2021 06:10 PM
    Hey everyone!

    I hope I'm posting in the right place. 
    I am trying to conduct a multiple imputation. I believe I have done everything correctly, but for whatever reason even though I requested for an iteration history to be written in the output option, when I run the analysis, the output shows that the requested iteration history output dataset can not be written. Do y'all know how to fix this?

    I thought it wouldn't be a problem, but when I run other analyses (i.e., descriptives), the pooled values don't show up. And the only thing I can think the problem is, is the iteration history.

    Any help would be appreciated, thank you!
    Allie =)

    ------------------------------
    Allie Dluzniewski
    ------------------------------

    #SPSSStatistics


  • 2.  RE: Iteration history for multiple imputation

    Community Leadership
    Posted Tue March 16, 2021 06:22 PM
    Thanks Allie! I've reached out to some of the experts on the IBM team who may be able to take a look at your question and help. 

    @Mathew Mathew
    @Sajan Kuttappa
    @Brent Gardiner
    @Ravinder Singh​​​​​

    ------------------------------
    Christina Howell
    ------------------------------



  • 3.  RE: Iteration history for multiple imputation

    Posted Fri September 24, 2021 11:07 AM
    Hi Allie,
    Have you got the answer? I am facing the same problem here.

    Sumi

    ------------------------------
    Sumina Shrestha
    ------------------------------



  • 4.  RE: Iteration history for multiple imputation

    IBM Champion
    Posted Fri September 24, 2021 12:14 PM
    If there was a problem with writing the imputation dataset, that might have stopped the procedure.

    If you chose to write new datasets with the log and imputed data, be sure that the datasets are declared ahead of the MULTIPLE IMPUTATION command using DATASET DECLARE.

    If they are not declared, they would be treated as files, which might not be valid filespecs.

    ------------------------------
    Jon Peck
    ------------------------------



  • 5.  RE: Iteration history for multiple imputation

    Posted Wed September 29, 2021 07:50 AM
    Hi Allie, 
    Check if the following link helps:
    https://www.ibm.com/docs/en/spss-statistics/24.0.0?topic=imputation-outfile-subcommand-multiple-command

    The page says two important things:
    1. The FCSITERATIONS keyword specifies an optional dataset that contains iteration history data for FCS imputation.
    2.
    The FCSITERATIONS keyword is ignored if the monotone method is used. A warning is shown and the dataset is not generated if all imputed variables are categorical.


    ------------------------------
    Vandit Desai
    ------------------------------