SPSS Statistics

 View Only
  • 1.  Comparing factor means

    Posted Tue August 15, 2023 11:43 AM

    Hello everyone

    I need help to compare the results of 2 groups (1= self identifies as sustainable consumer, 2= does not) regarding 5 different factors. Every factor contains around 5 variables that were measured on a 7-step Likert scale. I would like to provide an overview of the difference per factor, and not in detail per variable if possible. The aim of this is to show whether the results are significantly different for both groups.

    So far I have done an independent t-test, but I am unsure whether this is correct, as the factors are not continuous variables and are not normally distributed.

    I would be very thankful for your help!

    Kind regards

    Isabelle Koss

  • 2.  RE: Comparing factor means

    IBM Champion
    Posted Tue August 15, 2023 12:35 PM
    Why not use the Mann-Whitney test instead of the t test?


  • 3.  RE: Comparing factor means

    Posted Wed August 16, 2023 06:49 AM

    If you variables do constitute factor (in the factor analysis sense) you could generate the set of factors from the variables then do a multivarite analysis using your factors as dependent variables and the group variable as an independent factor (in the glm sense). If your sample size is reasonably large the resulting tests will approximately valid. If the sample size is small (<30) you could use a bootstrap approach to get valid test results.


    Graeme Byrne