• Home
  • Textbooks
  • Statistics for Marketing and Consumer Research
  • Hypothesis Testing

Statistics for Marketing and Consumer Research

Mario Mazzocchi

Chapter 6

Hypothesis Testing - all with Video Answers

Educators

Chapter Questions

04:00

Problem 1

Open the EFS data-set:
a. Using a one-mean $t$-test, test the hypothesis that the average number of cars or vans per household in the UK (variable a124) is exactly one
b. Is the distribution of the mean of the above variable normal? Plot the histogram and the normal curve, then apply a Kolmogorov-Smirnov test for normality
c. Test again the hypothesis of one car per household using a non-parametric test. Do results differ? Why?
d. Compare mean expenditure in restaurants and hotels (variable p611) between the lowest income quartile and the highest income quartile (as classified by variable INCRANGE). Are means significantly different? Does the assumption of equality of variance influence the results?

Ashley High
Ashley High
Numerade Educator
01:14

Problem 2

Open the Trust data-set:
a. Use the Chi-square test to check whether the sample is equally distributed across different categories of town size (q63)
b. 'About $30 \%$ of consumers shop in butcher's shop' - Test this hypothesis using the Binomial test on question q8d
c. Consider the general attitude toward chicken (q9) - Using the Runs test, check whether the sample is sorted in a random way
d. Consider that the data-set is sorted according to countries. Now select consumers from the UK only $(q 64=1)$ and repeat the Runs test. Do results differ? Why?

Akhil Choudhary
Akhil Choudhary
Numerade Educator
View

Problem 3

Open the Trust data-set:
a. Is expenditure in chicken (q5) normally distributed? Use the appropriate test and draw a histogram
b. Is the frequency of purchases of frozen chicken (q2d) compatible with a Poisson distribution?
c. Using rank tests, check whether those in the lowest income range like chicken (q9) more than those in the highest income range
d. Compute the estimated variance of chicken expenditure (q5) for female and male reference persons (q60) and perform an F-test to check for their equality - Hint: the $95 \%$ critical values are $\mathrm{F}_{0.025}(351,145)=1.33$ and $\mathrm{F}_{0.975}(351,145)=0.77$
e. Compare mean expenditures for q5 using the 2-independent samples t-test and check the Levene's F-statistic. Does it lead to a different conclusion?

Shu Naito
Shu Naito
Numerade Educator