SPSS Statistics

 View Only
Expand all | Collapse all

Multiple corrections for non-parametric t-test

  • 1.  Multiple corrections for non-parametric t-test

    Posted Tue January 17, 2023 07:59 AM
    Hey (:

    I don't have a lot experience with statistics or even SPSS, but I'm learning, I want to compare two separate groups on several measures (ordinal + continuous). Therefore, I wanted check first whether the two groups differed with respect to age. So I used the Mann-Whitney U Test. And unfortunately, they did. Even after correcting for multiple comparisons using Benjamini-Hochberg correction. My plan was to use Mann-Whitney U Tests on all measures. Now I'm wondering what I should do. In ANCOVAS one could use age as a covariate and group as an independent variable. ANOVA's are for three or more groups. So, my question is whether there exists a t-test which looks out for possible age-effects. 

    Thanks for your help. (:

    ------------------------------
    Linda Eckert
    ------------------------------

    #SPSSStatistics


  • 2.  RE: Multiple corrections for non-parametric t-test

    Posted Tue January 17, 2023 01:40 PM
    Regarding this: "ANOVA's are for three or more groups. So, my question is whether there exists a t-test which looks out for possible age-effects."

    Actually, you can use analysis of variance when you have 2 or more groups, and, in your example, make it an analysis of covariance by adding age to it.

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

    Original Message