EXAMPLE 11.13 Predicting U.S. Box Office Revenue

movies

CASE 11.2 In Example 11.12, there are explanatory variables, and we have data on movies. The degrees of freedom for multiple regression are therefore

Statistical software output for this fitted model provides many details of the model’s fit and the significance of the independent variables. Figure 11.11 shows multiple regression outputs from Excel, Minitab, and JMP. You see that the regression equation is

and that the regression standard error is .

image
Figure 11.11: FIGURE 11.11 Multiple regression outputs from Excel, Minitab, and JMP, Examples 11.13, 11.14, and 11.15.

553

The outputs present the statistic for each regression coefficient and its two-sided -value. For example, the statistic for the coefficient of Opening is 6.20 with a very small -value. The data give strong evidence against the null hypothesis

that the population coefficient for opening-weekend revenue is zero. We would report this result as , , . The software also give the 95% confidence interval for the coefficient . It is (1.42, 2.80). The confidence interval does not include 0, consistent with the fact that the test rejects the null hypothesis at the 5% significance level.