Binary Logistic Regression with Multiple Data Points

#1
Does anybody know how to run a regression on binary outcome data, with categorical predictors, where multiple subjects each have multiple outcome data points? I know it would be a logistic regression, however I'm not sure how to factor in having multiple subjects, each with multiple outcome observations.
 

BGM

TS Contributor
#2
Depends on your parameterization
For a categorical variable, if it have c categories,
then you can factor it with a group of dummy variables with c-1 degree of freedom.

For example, you can use c-1 {0, 1} dummy variables
for the first (can be any one) category, all dummy variables = 0
and for the other c-1 dummy variables, you would set the corresponding dummy
variables to be 1 and other to be 0.