Statistics for Business: Decision Making and Analysis, 3rd Edition

Think About It

1. 23. No. Regression on a dummy variable is comparable to a two−sample t−test with the pooled variance estimate.

2. 25. Divide estimates and standard errors by 1.5, and sums of squares by 1.5². The F and t−statistics remain as shown.

3. 27. Statistically significant with p−value less than 0.0001.

4. 29. Least squares minimizes deviations from the fit in each category, and that happens if the fit within each is the mean in that category.

5. 31. No. An analysis of variance does not assume a relationship among the means. The numerical coding implies that the means line up with equal spacing.

6. 33. Cannot tell; rejecting H₀ does not imply a specific confidence interval excludes zero.

7. 35. The chance for a Type I error is larger than 0.05; set α = ...