SPSS Statistics

SPSS Statistics

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

 View Only

  • 1.  Comparing means for 5 or more different variables (i.e., more than 5 columns)

    Posted Mon June 21, 2021 12:18 PM
    Dear All,
    Good evening.
    I am a complete novice in SPSS. I am currently trying to compare the means of 5 or more variables (i.e., 5 columns in spss). I have been reading and watching the videos on youtube. But they all describe about comparing means of different groups (i.e., male or female of one variable/column). But, I am not trying to use independent and dependent variables. I just want to compare the means of more than three different variables (i.e., means of more than 3 columns in SPSS spreadsheet). Is there anyone who can please help on the steps to do this? My reading and watching videos is not helping.
    Will be grateful for your understanding help.
    Have a great day,
    WM

    ------------------------------
    Wise Mainga
    ------------------------------

    #SPSSStatistics


  • 2.  RE: Comparing means for 5 or more different variables (i.e., more than 5 columns)

    Posted Mon June 21, 2021 01:03 PM
    Most of the procedures in SPSS Statistics use a grouping variable or factor for comparisons.  You would probably want to use Analyze > Compare Means > One-way ANOVA since you have more than two groups,  But to do that, you need to restructure your data from "wide form" to "long form".  That will create a single variable from your five along with an index variable that can be used as the group factor.

    To do this, use Data > Restructure  and choose Restructure selected variables into cases.  The wizard will walk you through the steps.

    For this to make sense, though, the variables need to be measures of the same thing.  Otherwise, the mean comparisons would probably not make sense.

    --



    Original Message


  • 3.  RE: Comparing means for 5 or more different variables (i.e., more than 5 columns)

    Posted Mon June 21, 2021 03:56 PM
    Edited by System Admin Fri January 20, 2023 04:11 PM
    Hello. You say that you want to compare the means of different variables, so you are indeed interested in dependent measures, at least.

    You have no grouping variable(s), so it sounds to me like you want a multivariate analysis of variance. We use the GLM procedure for that.

    Let's say your variables are v1 v2 v2 v4 v5. The syntax is:

    GLM V1 V2 V3 V4 V5 /METHOD=SSTYPE(3) /INTERCEPT=INCLUDE /CRITERIA=ALPHA(.05).

    There are some subcommands you could use to compare specific combinations of means. You can consult the documentation for those details.

    ------------------------------
    Rick Marcantonio
    Quality Assurance
    IBM
    ------------------------------

    Original Message