# SPSS/compute expected utility/compute correlation matrix

Not what you're looking for?

Please see the attached file for full problem description.

---

1. A respondent was presented 18 hypothetical alternatives or profiles described on five attributes of carpet cleaner. The attributes are:

o package design (described on three levels)

o brand name (described on three levels)

o price (described on three levels)

o Good Housekeeping seal of approval (described on two levels)

o money-back guarantee (described on two levels)

Based on the attributes, 108 possible profiles could be created. This is given by 3 × 3 × 3 × 2 × 2 = 108. The respondent was given a particular set of 18 and was asked to rank the profiles from the most preferred (rank of 1) to the least preferred (rank of 18). The data is given in the table below. The first five columns represent the five attributes and the last column is a respondent's ranking.

Package

Design Brand

Name Price Good

Housekeeping

Seal Money

Back

Guarantee Respond

(Rank)

A K2R 1.19 No No 13

A Glory 1.39 No Yes 11

A Bissell 1.59 Yes No 17

B K2R 1.39 Yes Yes 2

B Glory 1.59 No No 14

B Bissell 1.19 No No 3

C K2R 1.59 No Yes 12

C Glory 1.19 Yes No 7

C Bissell 1.39 No No 9

A K2R 1.59 Yes No 18

A Glory 1.19 No Yes 8

A Bissell 1.39 No No 15

B K2R 1.19 No No 4

B Glory 1.39 Yes No 6

B Bissell 1.59 No Yes 5

C K2R 1.39 No No 10

C Glory 1.59 No No 16

C Bissell 1.19 Yes Yes 1

6. Code all five attributes (including price) using dummy variables. Enter the ranks and the dummy variables for each of the 18 profiles into an SPSS data set. Convert the rank orders so that the higher the number the more preferred the object. In the table above, the most preferred object is ranked 1. Simply subtract each rank given in the last column from 19 to get the converted ranks. Hint: model your matrix after page 5 in the conjoint handout.

7. Compute the correlation matrix of the dummy variables. What do you notice in the correlation matrix? What does this say about the particular set of 18 profiles that was presented to the respondent? If any of the off-diagonal elements are not zero, explain why.

8. Run a regression using the rank as the dependent variable and the dummy variables as the predictors (Disregard your suspicion that the dependent variable is ordinal -- as I said in class, the dependent variable for regressions must be interval scaled.). The regression coefficients of the dummy variables can be treated as utilities.

9. Compute the relative importance of each attribute. Hint: see page 6 in the conjoint lecture.

10. Calculate the predicted utilities for the following options:

1. Package A, K2R, $1.19, no Good Housekeeping seal, no money-back guarantee

2. Package C, Bissell, $1.19, no Good Housekeeping seal, money-back guarantee

3. Package B, Bissell, $1.59, Good Housekeeping seal, money-back guarantee

11. What is the highest predicted utility that can be obtained? What are the characteristics of that option?

##### Purchase this Solution

##### Solution Summary

1. A respondent was presented 18 hypothetical alternatives or profiles described on five attributes of carpet cleaner. The attributes are:

o package design (described on three levels)

o brand name (described on three levels)

o price (described on three levels)

o Good Housekeeping seal of approval (described on two levels)

o money-back guarantee (described on two levels)

Based on the attributes, 108 possible profiles could be created. This is given by 3 × 3 × 3 × 2 × 2 = 108. The respondent was given a particular set of 18 and was asked to rank the profiles from the most preferred (rank of 1) to the least preferred (rank of 18). The data is given in the table below. The first five columns represent the five attributes and the last column is a respondent's ranking.

Package

Design Brand

Name Price Good

Housekeeping

Seal Money

Back

Guarantee Respond

(Rank)

A K2R 1.19 No No 13

A Glory 1.39 No Yes 11

A Bissell 1.59 Yes No 17

B K2R 1.39 Yes Yes 2

B Glory 1.59 No No 14

B Bissell 1.19 No No 3

C K2R 1.59 No Yes 12

C Glory 1.19 Yes No 7

C Bissell 1.39 No No 9

A K2R 1.59 Yes No 18

A Glory 1.19 No Yes 8

A Bissell 1.39 No No 15

B K2R 1.19 No No 4

B Glory 1.39 Yes No 6

B Bissell 1.59 No Yes 5

C K2R 1.39 No No 10

C Glory 1.59 No No 16

C Bissell 1.19 Yes Yes 1

6. Code all five attributes (including price) using dummy variables. Enter the ranks and the dummy variables for each of the 18 profiles into an SPSS data set. Convert the rank orders so that the higher the number the more preferred the object. In the table above, the most preferred object is ranked 1. Simply subtract each rank given in the last column from 19 to get the converted ranks. Hint: model your matrix after page 5 in the conjoint handout.

7. Compute the correlation matrix of the dummy variables. What do you notice in the correlation matrix? What does this say about the particular set of 18 profiles that was presented to the respondent? If any of the off-diagonal elements are not zero, explain why.

8. Run a regression using the rank as the dependent variable and the dummy variables as the predictors (Disregard your suspicion that the dependent variable is ordinal -- as I said in class, the dependent variable for regressions must be interval scaled.). The regression coefficients of the dummy variables can be treated as utilities.

9. Compute the relative importance of each attribute. Hint: see page 6 in the conjoint lecture.

10. Calculate the predicted utilities for the following options:

1. Package A, K2R, $1.19, no Good Housekeeping seal, no money-back guarantee

2. Package C, Bissell, $1.19, no Good Housekeeping seal, money-back guarantee

3. Package B, Bissell, $1.59, Good Housekeeping seal, money-back guarantee

11. What is the highest predicted utility that can be obtained? What are the characteristics of that option?

##### Purchase this Solution

##### Free BrainMass Quizzes

##### Know Your Statistical Concepts

Each question is a choice-summary multiple choice question that presents you with a statistical concept and then 4 numbered statements. You must decide which (if any) of the numbered statements is/are true as they relate to the statistical concept.

##### Measures of Central Tendency

This quiz evaluates the students understanding of the measures of central tendency seen in statistics. This quiz is specifically designed to incorporate the measures of central tendency as they relate to psychological research.

##### Terms and Definitions for Statistics

This quiz covers basic terms and definitions of statistics.

##### Measures of Central Tendency

Tests knowledge of the three main measures of central tendency, including some simple calculation questions.