Business Application of Regression Model Equation

1. WordMath is a new type of software that allows users to do symbolic mathematicl analysis (such as calculus) from within word processing software. There are two versions of the software: (i) a basic version and (ii) an advanced version. The advanced version provides more functions than the basic version. Both versions cost the same to manufacture and market. These costs are $30/unit.

Available marketing research data give the following segment characteristics.

Scientists and Engineers in manufacturing firms: Segment size is 1.2 million. Their willingness-to-pay (WTP) for the basic version is $350 and for the advanced version the WTP is $400.

Consulting and design firms: Segment size if 900,000. Their WTP for basic version is $385 and or the advanced version is $440.

Assume that the willingness-to-pay valuations include the effects of competition; each consumer needs only one product; and all consumer buy the product that gives them the highest net value (WTP minus price paid).

(a) If WordMath can sell only one version of the product, which version should it sell and at what price?
(b) If WordMath can sell both products should it sell both products? If it sells both products what prices should it charge? Be clear about which segment will buy which product.

2. The following table contains data from a survey looking at the importance of various attributes of laptops.

Configuration Hard Disc Brand Price Processor Warranty (years) Rank
1 80 Dell 2100 AMD 1 13
2 80 Philips 2800 AMD 2 11
3 80 Lenovo 3500 Intel 1 17
4 140 Dell 2800 Intel 2 2
5 140 Philips 3500 AMD 1 14
6 140 Lenovo 2100 AMD 1 3
7 100 Dell 3500 AMD 2 12
8 100 Philips 2100 Intel 1 7
9 100 Lenovo 2800 AMD 1 9
10 80 Dell 3500 Intel 1 18
11 80 Philips 2100 AMD 2 8
12 80 Lenovo 2800 AMD 1 15
13 140 Dell 2100 AMD 1 4
14 140 Philips 2800 Intel 1 6
15 140 Lenovo 3500 AMD 2 5
16 100 Dell 2800 AMD 1 10
17 100 Philips 3500 AMD 1 16
18 100 Lenovo 2100 Intel 2 1

These numbers represent a representative customer's ranking of varios product configurations. For example, configuration 1 is a Dell laptopwith an 80G hard drive, priced at $2100, with an AMD processor, and a 1 year warranty. This laptop was rated 13 our of 18 conigurations. Note, a 1 represents the highest rated laptop - that is configuration 18 was rated highest.

Based on these data we have to figure out the relative importance of various attributes. In order to do this run the following regression:

Rank = a1 + b1*Brand + b2*HardDisc + b3*Price + b4*Processor + b5*Warranty.

In this regression, use dummy variables for all the variables.

a. What is more important - warranty or the processor inside the computer? Why?
b. What is the most highly rated brand name?
c. What is the Dell Brand worth (in dollars) in this market?
d. Run another regression with price as a continuous variable. Does this change Dell's $ value?