SPSS Statistics

SPSS Statistics

Your hub for statistical analysis, data management, and data documentation. Connect, learn, and share with your peers! 

 View Only
Back to discussions
Expand all | Collapse all

Issues with Missing Strata or Cluster Variables in SPSS Complex Samples Analysis

  • 1.  Issues with Missing Strata or Cluster Variables in SPSS Complex Samples Analysis

    Posted Tue July 16, 2024 09:29 AM

    Hello SPSS Forum Community,

    I am currently working on a project using SPSS for complex samples analysis, specifically using a two-stage sampling design with strata and cluster variables. Despite double-checking and ensuring that my variable names (V023 for strata and V021 for clusters) match exactly between my dataset and the plan file (CSA.csaplan), I continue to encounter the following warning:

    "One or more strata or cluster variables found in the sample file do not exist in the joint inclusion probabilities file. Execution of this command stops."

    I have verified:

    • Variable names are spelled correctly and match case sensitivity.
    • The plan file (CSA.csaplan) specifies STRATA=V023 and CLUSTER=V021 under the DESIGN section.
    • Syntax order is correct (running CSPLAN before analysis commands).

    Despite these efforts, I'm unable to proceed with my analysis. Could someone please advise on how to resolve this issue or suggest additional troubleshooting steps?

    Thank you in advance for your assistance!

    Best regards

    Saurav



    ------------------------------
    Saurav Thapa
    ------------------------------


  • 2.  RE: Issues with Missing Strata or Cluster Variables in SPSS Complex Samples Analysis

    Posted Wed July 17, 2024 07:17 AM
    Hi,
    You probably don't have the joint inclusion probabilities? It is rather uncommon to have them since they require that you draw each cluster independently (which is not the case with common linear systematic pps designs). 
    If you have specified stratified sampling without replacement on your design statement replace that with the with replacement option. (WR). 
    That is probably not your design, but works for variance estimation.
    Best,
    Jon



    Original Message