Group the count responses on UTIs and catheter blockages in the data set "intake" into three levels: no occurrence, only once, and more than once. Use these three-level outcomes to assess

(a) whether catheter blockage is associated with UTI;

(b) whether catheter blockage and UTI have the same marginal distribution;

(c) whether the joint distribution of catheter blockage and UTI is symmetric;

(d) agreement as measured by the kappa coefficient.