• Home
  • Textbooks
  • Statistics for Marketing and Consumer Research
  • Multidimensional Scaling

Statistics for Marketing and Consumer Research

Mario Mazzocchi

Chapter 13

Multidimensional Scaling - all with Video Answers

Educators

Chapter Questions

02:37

Problem 1

Open the MDS data-set
a. Using the ALSCAL program, assume that the data are metric (interval) and compute the distance matrix between pairs of sports for the 50 respondents with two-dimensional MDS
b. Which sports are the most similar according to consumer perceptions? Which ones are the most different?
c. Save the final co-ordinates and compute the bivariate correlations with the expert panèl evaluations
d. Try the three-dimensional configuration. Which solution works better according to the STRESS value and $\mathrm{R}^2$ ?
e. Repeat the analysis, this time looking for similarities between respondents for the ten sports on two-dimensions
f. How many clusters of consumers appear from the graph?
g. Run a principal component analysis on the 50 consumers (hint - transpose the data-set), extract the first two components and run a cluster analysis. Do results look similar to the ones of MDS?
h. Save the final co-ordinates and compute the bivariate correlations with the expert panel evaluations

James Kiss
James Kiss
Numerade Educator
02:16

Problem 2

Open the MDS data-set
a. Using the PROXSCAL program, assume that the data are ordinal and compute the distance matrix between pairs of sports for the 50 respondents with two-dimensional MDS
b. Which sports are the most similar according to consumer perceptions? Which ones are the most different?
c. What are the differences between this distance matrix and the one produced by ALSCAL (see exercise 1)?
d. Try and change the type of measure, the initial configuration (the latter from the OPTION menu) and the number of dimensions and find the best output according to the STRESS function
e. Repeat the analysis, this time looking for similarities between respondents for the ten sports on two-dimensions

James Kiss
James Kiss
Numerade Educator
01:01

Problem 3

Open the MDS data-set
a. Run the MDS UNFOLDING program (as in this chapter's example) using the following assumptions (OPTIONS menu):
i. Identity scaling model
ii. Proximities are dissimilarities
iii. No transformation of proximities
iv. No intercept
v. 'Correspondence' initial configuration
vi. Strength of penalty term: 0.3
vii. Range of penalty term: 2.0
b. Look at the output and comment on the results, comparing them with the ones in the chapter's example
c. Change the following options:
- Ordinal transformation of proximities
- Include an intercept
- 'Ross-Cliff' initial configuration
- Strength of penalty 0.1
- Range of penalty term 1.0
d. Compare the results with the previous ones. Try changing the options one by one to see which one has the largest impact on the results
e. Save the final co-ordinates in a separate file and compute the correlations with the expert panel evaluations

Dominador Tan
Dominador Tan
Numerade Educator