• Home
  • Textbooks
  • Statistics for Marketing and Consumer Research
  • Discriminant Analysis

Statistics for Marketing and Consumer Research

Mario Mazzocchi

Chapter 11

Discriminant Analysis - all with Video Answers

Educators

Chapter Questions

Problem 1

Open the EFS data-set
a. Using discriminant analysis, explore how household income (incanon) and age of the household reference person (p396p) explain discrimination between those with a home computer and those without (a1661)
b. How many discriminant functions are estimated?
c. Are the group significantly different in terms of the predictor values?
d. Which predictor is most relevant in explaining classification?
e. Does the assumption of equality of group covariance matrices hold?
f. What are the correct prediction rates?
g. Does allocation of all units to the modal category generate a better or worse prediction rate?

Check back soon!

Problem 2

Open the Trust data-set
a. Consider the allocation of respondents according to their job status (q54)
b. Using MDA, explore the role of the following predictors in explaining classification: age (q51), household size (q56), financial condition of the household (q61)
c. According to ANOVA tests, which predictors are statistically different across groups? Which predictor has the highest discriminatory power?
d. Which discriminant functions are significant according to the Chi-square test?
e. How much of the total variance does the first discriminant function explain?
f. Considering the structure matrix, try and label the discriminant functions
g. Compute the percentage of correctly predicted cases. Does allocation of all units to the modal category generate a better or worse prediction rate?

Check back soon!

Problem 3

Open the EFS data-set
a. Consider the allocation of sampling unit according to the tenure type (a121)
b. Using MDA, explore the role of the following predictors in explaining classification: age (p396), household size (a049), income (incanon), total housing expenditure (p604) and number of children (a019)
c. Which predictors are irrelevant in explaining classification?
d. Repeat the analysis using the stepwise selection method and compare the results

Check back soon!