I have a dataset which contains only categorical data i.e.
A,B,C,D (like factors) for each predictor. There are 10 predictors and the dependent variable is binary,
UPDATE: MY predictors are answers for multiple choice questions for a questionnaire. So each predictor only takes on categorical values, i.e.
X_1 can be
X_2 can be
Is it feasible to fit a logistic regression over this dataset? Ideally, if I can fit a logistic regression the data, I will then use it for prediction over a set of test data, which again contains only categorical data.
What are the pitfalls that I should look out for?