Summary 571a. What is the chance that a typical college student will respond to the
promotion? A typical doctor? A typical lawyer? With respect to which
group is the promotion most effective?
b. How might information, such as shown in the table, be useful to
marketing and advertising managers?- The following table (compiled from police reports) shows the record of
automobile accidents for three age groups over the last year in a five-
county region.
Number of Drivers Having:
No Accidents 1 Accident 2 or Accidents
Age 17–30 90,243 12,050 1,822
Age 31–55 243,125 21,443 2,822
Over 55 149,674 16,621 2,293a. An analyst points out that of 57,051 drivers involved in accidents last
year, drivers aged 31 to 55 accounted for 24,265 cases, or some 43
percent of the total—a far greater proportion than any other age
group. Should one conclude that this age group has the highest-risk
drivers?
b. Which age group has the worst accident record? The best? Explain.
c. A separate analysis shows that for drivers aged 35 to 45, the rate of
accidents (one or more per year) is 9.3 per 1,000 drivers. For drivers
aged 65 to 75, the rate is 8.4 per 1,000 drivers. However, most studies
show that members in the younger group are much safer drivers than
those in the older group. Why might a simple comparison of accident
rates per driver be misleading? What other important factor should be
taken into account?- Consider once again the decision to redesign an aircraft (Problem 4 in
Chapter 12).
a. Find the expected value of perfect information about the redesign
program. Calculate separately the expected value of perfect
information about the U.S. government’s decision.
b. Suppose that management of the consortium questions its engineers
about the success or failure of the redesign program prior to
committing to it. Management recognizes that its engineers are likely
to be biased in favor of the program. It judges that if the program
truly will succeed, the engineers will endorse it 90 percent of the time,
but even if the program will fail, they will endorse it 50 percent of the
time. What is the likelihood of success in light of an endorsement?
What if the engineers do not endorse the program?
c13TheValueofInformation.qxd 9/26/11 11:02 AM Page 571