Unitrust sales. A bank’s marketing manager is analyzing the monthly sales of their unitrust products. He feels that it is important to focus on making the brand and its product more distinct from that of the competitors. He collects data on whether or not the customers are going to purchase (Yes = 1, No = 0), the amount spent on advertising

($000), and the number of competing unitrust products from 105 licensed financial institutions. Here is the partial output from a logistic regression model of purchased unitrust products, amount spent on advertising these, and the competing products.

a) Write out the estimated regression equation.

b) Find the predicted monthly sales when $15,000 is spent on advertising, and there are 105 competing unitrust products in the market.

c) What is the predicted purchased probability?

d) Find the predicted monthly sales when $3,500 is spent on advertising, and there are 103 unitrust products in the market.

e) What is the associated predicted probability?

*40. Online versus offline shopping. Store shopping has recently seen a decline as e-commerce has been growing three times faster than a traditional retail store. An analysis was conducted to test the difference, if any, in the customers’

shopping behavior between online and off line shopping.

For this, 150 participants were selected to collect this information from by considering the type of shopping

(online = 1 or off line = 0) as the response variable, and the product price ($) and satisfaction rating regarding their purchase experience (%). The output from this logistic regression is given here:

a) Write out the estimated regression equation.

b) Is the satisfaction rating of the customers’ purchase experience statistically significant in predicting whether or not a purchase is made online? Explain.

c) Is the product price statistically significant in predicting whether or not a purchase is made online? Explain.