Chapter 7, Analysis of Variance Video Solutions, Statistics for Marketing and Consumer Research

Problem 1

Open the Trust data-set
a. Build a table showing average chicken consumption (q4kilos) for different economic conditions, self-assessed as in q61
b. Perform one-way ANOVA to test the null hypothesis of mean equality
c. Considering the five response categories of q61 in the original order, use linear contrasts to test the following hypotheses:
d. Are the above comparisons orthogonal?
e. Consider a third (non-orthogonal) comparison and report results considering the correct significance levels:
i. $\mu_1+\mu_2=\mu_5$
f. Report results of post-hoc tests
i. $\mu_1+\mu_2=\mu_3+\mu_4$
ii. $\mu_1=\mu_2+\mu_5$
d. Are the above comparisons orthogonal?

Victor Salazar

Numerade Educator

View

Problem 2

Open the EFS data-set. Using the univariate GLM, do the following steps in a two-way ANOVA:
a. Test the effects of income (incrange) and age of the reference person (ageband) on expenditure for transport (p607). Is the interaction of the two factors significant?
b. Compute and comment on the Helmert planned comparisons
c. Perform the Bonferroni, Sidak and Tukey post-hoc tests
d. Estimate the effect size (OPTIONS menu)
e. Estimate the power (OPTIONS menu)
f. Comment on the results

Victor Salazar

Numerade Educator

Problem 3

Open the EFS data-set:
a. Using the univariate GLM menu, run a two-way ANOVA to check the influence of sampling month (a055) and income bracket (incrange) on mean expenditures for books (c95111)
b. Repeat the analysis considering the sampling month and income range as a random rather than fixed factor. Do results change? Why?
c. Instead of using income as a categorical variable, include the metric (scale) measure of income (incanon) as a covariate and perform ANCOVA. Comment the results

Check back soon!