Table 7.14 is based on automobile accident records supplied by the state of Florida Department of Highway Safety and Motor Vehicles. Subjects were classified by whether they were wearing a seat belt, whether ejected, and whether killed.

a. Find a loglinear model that describes the data well. Interpret the associations.

b. Since the sample size is large, goodness-of-fit statistics are large unless the model fits very well. Calculate the dissimilarity index for the model you found in

(a) and interpret.

c. Conduct a Bayesian analysis of the homogeneous association model. Find and interpret posterior intervals for the conditional odds ratios. How do interpretations differ from ones you make with a frequentist analysis?