This is a statical question.
I have a data set that provides the number of people readmitted to associated hospitals within 30 days of discharge (a common efficacy measure). At immediate level of measurement I have categorical data:
Readmitted: Yes / No
I am confused about the implications for the unit of measure I am using: the number of of readmissions per zip code, over a one-year period of time.
Note that the unit of measure I am looking at I could 0 -> n readmissions (interval data).
Can the unit of measure be used to transform categoric data interval data? Are there special precautions in this situation (I intend to run one of the following ANCOVA, logistic regression, or decision tree)?