SPSS Statistics

SPSS Statistics

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

 View Only
Back to discussions
Expand all | Collapse all

Classification table in binary logistic regression predicting all cases as one category member?

  • 1.  Classification table in binary logistic regression predicting all cases as one category member?

    Posted Mon January 11, 2021 11:56 PM

    I am attempting to run a binary logistic regression using a number of categorical and continuous variables. However, every time I run the analysis the classification table predicts all cases to fall into one category of my outcome variable (e.g. males=1 and females=2, all predicted to be males). In this example, percentage correct for males is reported to be 100% and 0% for females. It does this no matter which binary variable I use as my dependent variable and no matter how they are coded (e.g. 1 and 2, 0 and 1).

    I'm new to logistic regression so any help would be much appreciated!

    LOGISTIC REGRESSION VARIABLES sex

    /METHOD=ENTER age canonpre

    /CONTRAST (canonpre)=Indicator

    /SAVE=PRED PGROUP COOK LEVER DFBETA RESID

    /CLASSPLOT

    /CASEWISE OUTLIER(2)

    /PRINT=GOODFIT ITER(1) CI(95)

    /CRITERIA=PIN(0.05) POUT(0.10) ITERATE(20) CUT(0.5).






    #SPSSStatistics
    #Support
    #SupportMigration


  • 2.  RE: Classification table in binary logistic regression predicting all cases as one category member?

    Posted Wed January 13, 2021 02:10 PM

    The coding doesn't matter, but if one dependent variable category is a lot bigger than the other, for example, this could happen. If the data are not very imbalanced, you might try different covariates to see the effect. Save the predicted probabilities and look at their distribution. Changing the threshold value for classification might be necessary.






    #SPSSStatistics
    #Support
    #SupportMigration


Related Content