SPSS Statistics

 View Only

  • 1.  Weighting Discrepancy with reporting correct Unweighted Base

    Posted 25 days ago

    I am having an issue with SPSS not correctly reporting the correct unweighted base for a weighted table. Here are the unweighted numbers. Any one else having this issue? 

    When I weight by the age cohorts the total column reflects the correct weighted mean but the incorrect unweighted base.  For example, the base for the total column should be 61.  Is there a different custom table statistic that I should use to report the unweighted base? I also don't understand why the bases for the other age columns are off. I have tried the other count /base statistics but none matches the correct unweighted base.  Unweighted counts reflect the counts for the full data set and not the reduced base for this question (which is what should show)

    Thanks in advance. Hoping there is a simple fix. 



    ------------------------------
    Ratna Wynn
    ------------------------------


  • 2.  RE: Weighting Discrepancy with reporting correct Unweighted Base

    Posted 25 days ago
    The tables in the email are too small for me to read, but there are six different count statistics, ignoring the mult response statistics.
    [COUNT UCOUNT ECOUNT VALIDN EVALIDN UVALIDN]

    Which are you using?  (If in doubt, put them all in and see if any give the result you expect).  You can write the list as above without all the formatting and labelling.

    If that doesn't solve the problem and you what to send me data and syntax, I can take a look.  (jkpeck@gmail.com)




    --



    Original Message


  • 3.  RE: Weighting Discrepancy with reporting correct Unweighted Base

    Posted 25 days ago

    Hi Jon - I will shoot you an email with the data and syntax. I have tried Valid N/Adjusted N/Unweighted Count, Total N, Adjusted Total N. None match. Thanks for the help



    ------------------------------
    Ratna Wynn
    ------------------------------

    Original Message