Exercise 2.2

(a) CEO salary

Regression, since the variable we are interested in, CEO salary, is quantitative. Since we want to understand the factors affecting CEO salary we are most interested in inference. The number of predictors is 3, and the sample size is 500 (p=3, n=500).

(b) New product launch: success or failure?

Classification, because we are studying the outcome of the launch of a new product, which we are representing in one of two categories: success or failure. In this case we wish to know whether a new launch will be a success or a failure more than understand why, so we are more interested in prediction. The number of predictors is 13, and the sample size is 20 (p=13, n=20).

(c) Change in the US dollar

Regression, since the variable we want to predict is a real number, and we are more interested in prediction. The number of predictors is 3, and the sample size is 52 (p=3, n=52).