Among 29,531 drivers the following numbers of accidents per driver were observed (in Connecticut, 1931-1936):
The overall average is 0.24 accidents per operator. If the accident rate is the same for all operators, the number of accidents for any driver has a Poisson distribution. Test the hypothesis that there is no accident "proneness" in certain drivers by comparing these numbers with a Poisson distribution. Use α = 10 percent. (Proneness includes here, of course, the factor that a driver simply may drive many more than the average number of miles and hence be exposed to more accident possibilities.)