Validation for Binary Logit Regression for Categorical Data


New Member

I'd like to know how to validate my fitted binary logit regression model for categorical data.

For example:
Predictors : A - Age (5 categories) , W - Weight (2 categories)
Response: 0 or 1
Fitted model:
logit(success/(trials-success) = function (A,W)

Since I have categorized a few thousands of individual data into 10 categories ( 5 groups of Age x 2 groups of Weight), how can I check the accuracy of my model using the data of individual data which I set aside for validation purpose?

Many thanks,