Consider how this analysis would be done differently if we were interested in whether cigarettes per day were associated with diabetes. 1. How is the variable CIGPDAY set up (what is its level of measurement)? 2. Would you use a different statistical test? If so, which one? diabetes is a binary variable