SPSS Statistics

 View Only

  • 1.  Using Multiple-response Answers in Chi-square test

    Posted Mon August 23, 2021 06:15 PM
    Hello, 

    I am currently trying to use SPSS to run a Chi-squared test on categorical multiple response questions and another categorical question. To do this, I am going to Analyze, tables, define multiple response sets, where I am selecting the needed variables and naming the set. When I run the chi-square test, I am getting a warning, saying
    I am not sure what is wrong with the way I set up the table. I had put count in the summary statistics. 

    Another thing I am not sure about is when I am defining the variable set, I am seeing "Two or more set variables use different labels for the counted value, or some variables for that value and some do not. Apply the value label from the first variable that has a label to all other set variables, now?" I'm not sure what this means. All of the responses are ones or zeros. 

    Any help is appreciated.

    ------------------------------
    Nathanael Sovitzky
    ------------------------------

    #SPSSStatistics


  • 2.  RE: Using Multiple-response Answers in Chi-square test

    IBM Champion
    Posted Mon August 23, 2021 06:29 PM
    Edited by System Fri January 20, 2023 04:38 PM
    I would have to see the table to figure this out.
    The warning message you are getting is because the 0 and 1 values are labelled differently for different variables.  That may be okay, but it might indicate a wrong set definition.  If one variable labels 0/1 yes/no and another labels 0/1 red/blue, they are probably not appropriate in the same set.

    Also, you referred to the chi-squared test, but you  might want a column proportions test.
    --



    Original Message


  • 3.  RE: Using Multiple-response Answers in Chi-square test

    Posted Tue August 24, 2021 09:53 AM
    Try this if it works:
    Go to transform >> Recode into new variable >> Select your first categorical variable, give a new name to the new variable, assign values to each category (click "old and new values"), 
    Do this for the second variable too. 
    Run the Chi-square test on newly created variables.

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

    Original Message