SPSS Statistics

SPSS Statistics

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

 View Only
  • 1.  intercept in MANOVA

    Posted Wed January 04, 2023 04:40 PM
    I am running a MANOVA to compare two groups on 17 different dependent variables. In my output below, I see the variable that shows the group effect (for NeuroDX_record_change_1). But what does the intercept in this model mean? This is also statistically significant, but it is unclear to me what this is testing. Any thoughts? Thanks, Kevin

    Multivariate Testsa

    Effect

    Value

    F

    Hypothesis df

    Error df

    Sig.

    Intercept

    Pillai's Trace

    .745

    101.294b

    6.000

    208.000

    <.001

    Wilks' Lambda

    .255

    101.294b

    6.000

    208.000

    <.001

    Hotelling's Trace

    2.922

    101.294b

    6.000

    208.000

    <.001

    Roy's Largest Root

    2.922

    101.294b

    6.000

    208.000

    <.001

    NeuroDX_record_change_1

    Pillai's Trace

    .182

    7.689b

    6.000

    208.000

    <.001

    Wilks' Lambda

    .818

    7.689b

    6.000

    208.000

    <.001

    Hotelling's Trace

    .222

    7.689b

    6.000

    208.000

    <.001

    Roy's Largest Root

    .222

    7.689b

    6.000

    208.000

    <.001

    a. Design: Intercept + NeuroDX_record_change_1

    b. Exact statistic



    ------------------------------
    Kevin Duff
    ------------------------------

    #SPSSStatistics


  • 2.  RE: intercept in MANOVA

    Posted Thu January 05, 2023 11:06 AM
    Hi. As with regression, tests of the intercept test the hypothesis that an obtained value (e.g., a single mean for univariate ANOVA, a combination of means in MANOVA) differs from 0. This test only makes sense to interpret if 0 has a meaning for your dependent measures; that is, if a subject can actually have the value 0.

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



  • 3.  RE: intercept in MANOVA

    Posted Thu January 05, 2023 11:53 AM
    Rick - Thanks for the quick and clear response to my question. The dependent variable is a z-score, so 0 definitely means something. Kevin