Errors An auto parts company advertises that its special oil additive will make the engine “run smoother, cleaner, longer, with fewer repairs.” An independent laboratory decides to test part of this claim. It arranges to have a taxicab company’s fleet of cars use the additive, and follows them to see how many need repairs. At the end of a year the laboratory will compare the percentage of cars that need repairs to the historical repair percentage.

1. What kind of a study is this?

2. Will they do a one-tailed or a two-tailed test?

3. Explain in this context what a Type I error would be.

4. Explain in this context what a Type II error would be.

5. Which type of error would the additive manufacturer consider more serious?

6. If the cabs with the additive do indeed run significantly better, can the company conclude it is an effect of the additive? Can they generalize this result and recommend the additive for all cars? Explain.

7. How might you improve the study?