SPSS Statistics

Expand all | Collapse all

predicted in genlinmixed

  • 1.  predicted in genlinmixed

    Posted Fri September 18, 2020 12:34 PM
    I run a genlinmixed in spss 26 (i added fixpack1, but this did not help) with weeks nested in ID:

    GENLINMIXED /DATA_STRUCTURE SUBJECTS= id*weeks

    /FIELDS TARGET= my_DV TRIALS=NONE OFFSET=NONE

    /TARGET_OPTIONS DISTRIBUTION=NORMAL LINK=IDENTITY

    /FIXED EFFECTS= my_covariate

    USE_INTERCEPT=TRUE

    /RANDOM EFFECTS= my_covariate

    USE_INTERCEPT=TRUE SUBJECTS=id COVARIANCE_TYPE=VARIANCE_COMPONENTS

    of around 460 subjects (ID). Around 400 are fine (predicted and observed do fit), but some are ugly (e.g. observed incline, but predicted decline and vice versa).

    Is that a general = known problem? can someone help?

    ------------------------------
    Micha
    ------------------------------


  • 2.  RE: predicted in genlinmixed

    Posted Fri September 18, 2020 12:54 PM
    That sounds like a problem with your model, not the procedure.

    ------------------------------
    Jon Peck
    ------------------------------



  • 3.  RE: predicted in genlinmixed

    Posted Mon September 21, 2020 11:07 AM
    Thanks

    ------------------------------
    sandy s
    ------------------------------