QUESTION 1: RESEARCH DESIGN, MEASUREMENT, SCALING & SAMPLING (total: 21 points)
An advertising agency wants to know who is watching TV when so they know when to air certain
commercials. The agency is interested in both the amount of time someone watches and the type of
shows someone watches. The agency wants to get information on age, gender and income.
Additionally the advertising agency wants to get information on whether someone lives in the city or
in more rural areas. The reason for the latter variable is that they think that people in rural areas watch
more television, because they have less alternative activities (e.g. going out).
1a. What type of research do you advise the advertising agency to conduct: exploratory research,
descriptive research, or causal research? Please motivate your answer (5 points)
Answer:
Motivation:
1b. The advertising agency decides to develop a questionnaire. To ask the age of the respondents the
agency uses the following question and scale:
How old are you?: <18 18-25 25-40 40-65 >65
Explain that the scale is ordinal. And give an example of an interval and ratio version of this question.
(4 points)
Motivation ordinal:
Interval version:
Ratio version:
1c. To ask about the type of shows the respondents watch the agency includes the following question:
I mainly watch television for news and knowledge enhancement:
Strongly disagree Disagree Neither agree/disagree Agree Strongly Agree
Is this a comparative or non-comparative scale? What is the measurement level of this scale? Give
arguments. Additionally, indicate two problems of the formulation of the question. How would you
improve this? (6 points)
Type of scale:
Measurement level + motivation:
What is wrong with question formulation?:
How would you improve the stem:
1d. The agency gets the opportunity to take a sample of actual household viewing data from a
television provider. The provider has lists of households per region. The regions are formed based on
the television towers that are used to send the signal. The geographical size of each region differs as
the coverage of a television tower is based on the number of households that it caters to. Explain how
cluster and stratified sampling would work based on the regions. Based on the belief that the
advertising agency has that people in the cities have different viewing patterns than in rural areas,
which one would you recommend. Motivate your answer. (6 points)
Cluster sampling:
Stratified sampling:
Preferred sampling + motivation:
QUESTION 2: HYPOTHESIS TESTING (total: 13 points)
A bank asks a panel of respondents across three countries for their motivation to use the option of
mobile banking. The questionnaire consists of several multi-item scales. From these the bank creates
three new variables that are called after their underlying construct: IntrinsicMotivation,
PerceivedHurdles and ExternalMotivation
2a. The bank wants to know if respondents of the three countries (UK, Sweden and Spain) have
different intrinsic motivation. The bank does an One-Way ANOVA. Please interpret the output below.
Assume that α = 0.05. Be as precise as possible. (5 points)
Conclusion + interpretation on ANOVA:
2b.The bank also thinks that males care more about what others think, and that this affects their
external motivation. Formulate the null and alternative hypothesis. Please interpret the output below.
In your answer, you should address the means of both groups, the Levene’s test, and the t-test.
Assume that α = 0.05. Be as precise as possible. (8 points)
Formulate the H0 and H1:
Conclusion + interpretation on means of both groups:
Conclusion + interpretation on Levene’s test:
Conclusion + interpretation on T-test: