OE106 Research Skills for Marketeers
Name: Dani Dorenbos
Student number: 671429
Class: 3B
Teacher: Hugo Boons
Deadline: 17 April 2023
,OE106 Research Skills for Marketeers
Dani Dorenbos, 671429, 3B-1
Hugo Boons
Inholland Alkmaar
,Preface
My name is Dani Dorenbos with student number 671429 from class 3B. For my second semester in
my third year, I got challenged with the assignment of SPSS for the course OE106. I was challenged to
make multiple tests with SPSS. These tests where:
General data mining
Logistic regression
Hierarchical cluster analyses
Linear regression
I had the hardest time with the Linear regression because it was really time consuming and took a lot
of my concentration. I was totally 4 whole days busy with this test.
I want to thank Hugo Boons for giving me this challenge and I have learned lots of the usage of SPSS
and can say for my self that I am getting good in SPSS.
, Inhoud
A. General data mining..........................................................................................................................6
A1. Do male customers generate a higher revenue (total amount) than female customers, on?.......6
A2. Is there a difference in average age depending on the size of the household a customer is........7
part of?...............................................................................................................................................7
A3. Can the revenue be regarded as normally distributed?................................................................8
A4. The average number of times that meal 6 was ordered, used to be 2.5 before a small change in
the recipe was made at the start of the past year. Has this number significantly changed the past
year? (In what direction?)...................................................................................................................9
A5. Is the distribution of age classes the same for males as for females, or is one of the two on a
significantly older level? (This is not about the exact average age.).................................................10
A6. Is there a difference in overall ranking level between household types? (The variable ‘ranking’
is not metric here.)...........................................................................................................................12
A7. Is there a relationship between total number of meals ordered and age? What kind of
relationship? How strong is it?.........................................................................................................12
A8. Is there a relationship between urbanization and age class? If so, what kind of relationship,
how strong is it? (Again, this is not about the exact age.).................................................................14
A9. Is there a relationship between household type and education? If so, try to describe it...........15
A10. Is the distribution of customers over the regions (provinces) equal to that of the general
population of the Netherlands? Check with percentages from the Bureau of Statistics CBS
mentioned above. [Hint: enter percentages without decimal comma or dot.] If not, which provinces
stand out with relatively few or many customers?...........................................................................17
B. Logistic regression............................................................................................................................19
B1. Check the significance of the relationship of all variables to be used with ‘risotto’ on a one-to-
one basis. Do not use variables which are derived from others (total, total amount, and age class).
[8 pts]...............................................................................................................................................19
B2. Devise a model, based on the variables measured here, to predict whether someone will order
the new meal. (Strictly speaking devise, a model that assigns a probability of someone saying that
they will order the meal to the values of a set a suitable variable measured for that person). Show
the intermediary stages. Make clear which variables you enter as ‘categorical’. 2...........................26
[12 pts].............................................................................................................................................26
B3. Evaluate the final model. [5 pts] This part has a maximum score of 25 points...........................30
C. Hierarchical cluster analysis..............................................................................................................31
C1. Show the dendrogram. Draw a line where you virtually ‘cut’ it to obtain four clusters. Mention
the variables used, cluster method (what distance is measured) and the measure, used. [7 pts.]...31
C2. (a) How many cases are attributed to each of the clusters? (b) Show the distribution of the
variables over the clusters (both the variables used in clustering and those not used in clustering.)
[10 pts.]............................................................................................................................................34
C3. Characterize the clusters with help of (2). [8 pts.]......................................................................56
Name: Dani Dorenbos
Student number: 671429
Class: 3B
Teacher: Hugo Boons
Deadline: 17 April 2023
,OE106 Research Skills for Marketeers
Dani Dorenbos, 671429, 3B-1
Hugo Boons
Inholland Alkmaar
,Preface
My name is Dani Dorenbos with student number 671429 from class 3B. For my second semester in
my third year, I got challenged with the assignment of SPSS for the course OE106. I was challenged to
make multiple tests with SPSS. These tests where:
General data mining
Logistic regression
Hierarchical cluster analyses
Linear regression
I had the hardest time with the Linear regression because it was really time consuming and took a lot
of my concentration. I was totally 4 whole days busy with this test.
I want to thank Hugo Boons for giving me this challenge and I have learned lots of the usage of SPSS
and can say for my self that I am getting good in SPSS.
, Inhoud
A. General data mining..........................................................................................................................6
A1. Do male customers generate a higher revenue (total amount) than female customers, on?.......6
A2. Is there a difference in average age depending on the size of the household a customer is........7
part of?...............................................................................................................................................7
A3. Can the revenue be regarded as normally distributed?................................................................8
A4. The average number of times that meal 6 was ordered, used to be 2.5 before a small change in
the recipe was made at the start of the past year. Has this number significantly changed the past
year? (In what direction?)...................................................................................................................9
A5. Is the distribution of age classes the same for males as for females, or is one of the two on a
significantly older level? (This is not about the exact average age.).................................................10
A6. Is there a difference in overall ranking level between household types? (The variable ‘ranking’
is not metric here.)...........................................................................................................................12
A7. Is there a relationship between total number of meals ordered and age? What kind of
relationship? How strong is it?.........................................................................................................12
A8. Is there a relationship between urbanization and age class? If so, what kind of relationship,
how strong is it? (Again, this is not about the exact age.).................................................................14
A9. Is there a relationship between household type and education? If so, try to describe it...........15
A10. Is the distribution of customers over the regions (provinces) equal to that of the general
population of the Netherlands? Check with percentages from the Bureau of Statistics CBS
mentioned above. [Hint: enter percentages without decimal comma or dot.] If not, which provinces
stand out with relatively few or many customers?...........................................................................17
B. Logistic regression............................................................................................................................19
B1. Check the significance of the relationship of all variables to be used with ‘risotto’ on a one-to-
one basis. Do not use variables which are derived from others (total, total amount, and age class).
[8 pts]...............................................................................................................................................19
B2. Devise a model, based on the variables measured here, to predict whether someone will order
the new meal. (Strictly speaking devise, a model that assigns a probability of someone saying that
they will order the meal to the values of a set a suitable variable measured for that person). Show
the intermediary stages. Make clear which variables you enter as ‘categorical’. 2...........................26
[12 pts].............................................................................................................................................26
B3. Evaluate the final model. [5 pts] This part has a maximum score of 25 points...........................30
C. Hierarchical cluster analysis..............................................................................................................31
C1. Show the dendrogram. Draw a line where you virtually ‘cut’ it to obtain four clusters. Mention
the variables used, cluster method (what distance is measured) and the measure, used. [7 pts.]...31
C2. (a) How many cases are attributed to each of the clusters? (b) Show the distribution of the
variables over the clusters (both the variables used in clustering and those not used in clustering.)
[10 pts.]............................................................................................................................................34
C3. Characterize the clusters with help of (2). [8 pts.]......................................................................56
