Name: Student number:
Exam Marketing Research Methods
Wednesday, November 7 2012
TEACHER
The exam contains XX pages.
The exam consists of 4 parts with several sub-questions.
Please answer each question in the corresponding answer box.
You have 3 hours to complete the exam (14.00-17.00 hrs).
Make sure that you write your name and student number on each page!
Please answer in English (Dutch is not allowed by the faculty).
The grades for the exam will be published on Nestor.
We wish you lots of success!!!
Maximum number of points per question:
Part 1: Regression, Third variables & Binary data 28 points
Part 2: Conjoint analysis 22 points
Part 3: Factor analysis 31 points
Part 4: ClusterAnalysis 19 points
1
, Name: Student number:
Part 1. Regression, Third variables & Binary data (28 points)
MarkS Consulting is a marketing research agency providing insights to a wide range of companies.
One of its services is a quarterly journal in which it publishes the results of state of the art analyses
on a wide variety of managerially relevant marketing topics.
For the next issue of the journal, MarkS the expenditures on advertising. Dependent variable was the
amount a company spent in a certain week W on advertising. Independent variables were the sales in
the previous week, the advertising investments of the company in previous week, and the advertising
investments made by competitors in previous week. All variables were expressed in logarithms, so
parameters can be interpreted as elasticities.
a) The level of advertising expenditures can be quite different in different product classes. The
researcher therefore decided to divide the companies into four classes (drinks, food, household
care and personal care), each represented by a dummy variable. Subsequently, he included
these four dummy variables in the model as explanatory (independent) variables. This way, he
could also get better insights on what product class is more/less advertising intensive. Do you
agree with this way of analyzing things? Provide arguments for your choice. (6 points)
No, this is not a good way to analyze things. The model will certainly crash because of perfect
multicollinearity. When using M category dummies, you should only include M-1 of them in the
model, and use one category as a benchmark category (in case of indicator coding like here).
2 points for the decision, 4 for the reasoning
In another article for the new issue, MarkS analyses the crisis on the housing market. In this article,
the research team tried to find out what the drivers are behind prices of houses. The dependent
variable hence is the price of the house, and the independent variables that were included in the
study are
- the number of rooms,
- the size of the house in square meters,
- the number of bathrooms,
- the age of the house,
- a dummy variable indicating whether it is a house (1) or an apartment (0)
- the state of the economy, expressed by the Consumer Confidence Index.
2
Exam Marketing Research Methods
Wednesday, November 7 2012
TEACHER
The exam contains XX pages.
The exam consists of 4 parts with several sub-questions.
Please answer each question in the corresponding answer box.
You have 3 hours to complete the exam (14.00-17.00 hrs).
Make sure that you write your name and student number on each page!
Please answer in English (Dutch is not allowed by the faculty).
The grades for the exam will be published on Nestor.
We wish you lots of success!!!
Maximum number of points per question:
Part 1: Regression, Third variables & Binary data 28 points
Part 2: Conjoint analysis 22 points
Part 3: Factor analysis 31 points
Part 4: ClusterAnalysis 19 points
1
, Name: Student number:
Part 1. Regression, Third variables & Binary data (28 points)
MarkS Consulting is a marketing research agency providing insights to a wide range of companies.
One of its services is a quarterly journal in which it publishes the results of state of the art analyses
on a wide variety of managerially relevant marketing topics.
For the next issue of the journal, MarkS the expenditures on advertising. Dependent variable was the
amount a company spent in a certain week W on advertising. Independent variables were the sales in
the previous week, the advertising investments of the company in previous week, and the advertising
investments made by competitors in previous week. All variables were expressed in logarithms, so
parameters can be interpreted as elasticities.
a) The level of advertising expenditures can be quite different in different product classes. The
researcher therefore decided to divide the companies into four classes (drinks, food, household
care and personal care), each represented by a dummy variable. Subsequently, he included
these four dummy variables in the model as explanatory (independent) variables. This way, he
could also get better insights on what product class is more/less advertising intensive. Do you
agree with this way of analyzing things? Provide arguments for your choice. (6 points)
No, this is not a good way to analyze things. The model will certainly crash because of perfect
multicollinearity. When using M category dummies, you should only include M-1 of them in the
model, and use one category as a benchmark category (in case of indicator coding like here).
2 points for the decision, 4 for the reasoning
In another article for the new issue, MarkS analyses the crisis on the housing market. In this article,
the research team tried to find out what the drivers are behind prices of houses. The dependent
variable hence is the price of the house, and the independent variables that were included in the
study are
- the number of rooms,
- the size of the house in square meters,
- the number of bathrooms,
- the age of the house,
- a dummy variable indicating whether it is a house (1) or an apartment (0)
- the state of the economy, expressed by the Consumer Confidence Index.
2
