1a. Rank the networks in terms of average ratings for TV movies during 1992.

1b. On average, how much higher are the ratings for the leading network than the ratings for the second-highest network?

2a. In 1992, what were the average ratings for fact-based movies?

2b. In 1992, what were the average ratings for fictional movies?

3. Consider Regression 2. Is the difference between the ratings for fact-based and fictional movies statistically significant? Explain.

4. Compare Regression 2 and Regression 3. Do the regressions suggest that, on average,

a. a fact-based movie has fewer stars than a fictional movie;

b. a fact-based movie has more stars than a fictional movie;

c. a fact-based movie has just as many stars as a fictional movie;

d. cannot be determined.

Choose one and explain.

For the next two questions, consider Regression 5.

5. On Sunday nights, CBC usually presents "Josette and Yvette" at 8:00 p.m., followed by the Sunday night movie at 9:00 p.m. Typical ratings for "Josette and Yvette" are 17.5. This week, Warrington is considering replacing "Josette and Yvette" with a live rock concert that is expected to gamer a rating of 20 points. What is the expected change in ratings for the Sunday night movie?

6a Warrington fears that a movie with high expected ratings might provoke the other networks to schedule better programming against CBC. Suppose that in response to CBC's programming, both ABN and BBS schedule different programs, each of which is expected to rate 2 rating points higher. What is the expected impact on the ratings of CBC's TV movie?

6b Oskar Morgenstern, a CBC network executive, believes that network programming does not affect the size of the total television audience in a given time slot. Instead, he believes that a network's programming only determines the network's percentage share of the total audience. Does Regression 5 support Morgenstem's position? Explain.

7. Warrington believes that movies with stars tend to be shown in favorable time slots (e.g., good months, good days of the week, and following highly rated programs).

a. Are the regressions consistent with her beliefs? Explain.

b. Warrington is planning to add a fictional movie to the programming schedule. She must decide whether or not to use a star. What is the difference in expected ratings between using a star and not using a star?

8. The conventional industry wisdom is that fact-based movies have higher ratings than movies based on fictional stories. Do the regressions support or contradict this view?

9. Warrington wants to put the TV movie in the best possible slot so as to help ensure high ratings. She has 3 slots available:

if Warrington wants to maximize the chance of high ratings, when should she schedule the TV movie?

10. Warrington is unsure of which TV movie to schedule. Due to the limited budget for a TV movie, CBC can choose either a fictional movie with a star or a fact-based movie without a star. Both movies are identical in all other respects. Assuming she wishes to maximize ratings, which movie should Warrington choose?

For the next two questions, assume that a normal distribution with mean M and standard deviation s, can be approximated with the following discrete 5-point distribution:

Probability Value

.20                m -1.3s

.20                 m - 0.5s

.20

.20              m + 0.5s

.20                                m -r 1.3s .

Thus, each point gets the same probability, .20.

11. Suppose that Warrington has scheduled a fact-based movie without a star for a Monday time slot in March (again, following a show that typically receives ratings of 13.0). Should Warrington accept Harsanyi Electric's offer or accept the fixed fee of $5,000,000?

12. Suppose that, prior to accepting or rejecting Harsanyi Electric's offer, Warrington could purchase a regression that would tell with virtual certainty what the Nielsen rating of the proposed movie would be. What is the most that Warrington would be willing to pay for such a regression?