Validate Binary Logistic Regression model in SPSS Statistics?

I use 60% of my data with SPSS Statistics to build a Binary Logistic Regression model to predict a disease outcome (with 0 normal, 1 disease),

then, I'd like to validate the Binary Logistic Regression model with the reste of data (40%) by using Scoring Wizard (Utilities of SPSS, Scoring Wizard to apply a model created with one dataset to another dataset and generate scores, such as the predicted value and/or predicted probability of an outcome of interest),

after that, I'd like to show the ROC curve of the outcome of the the validation results (40% data)

My question is which value I need to use to show the ROC curve of the validation results (40% data) in the output of Scoring Wizard, Predicted value, Probability of predicted value or Probability of selected value ?

Thanks

More Xiangping Lin's questions See All
Similar questions and discussions