19-4-2022
Research
OE106
Timo Rouwenhorst
645803
Hogeschool Inholland te Haarlem
Business Studies
Timo Rouwenhorst, 645803, HABSMAVT3B
,Inhoud
Part A: Exploring the data.......................................................................................................................2
Part B: Logistic regression.....................................................................................................................10
Part C. Linear regression.......................................................................................................................18
Part D: Hierarchical cluster analysis......................................................................................................21
1 Timo Rouwenhorst, 645803
, Part A: Exploring the data
1. Do the customers who receive a newsletter generate a higher revenue than the others, on
average?
I used the Independent Sample T-Test, because there is a group who receives a newsletter
and a group who do not.
H0: Do the customers who receive a newsletter generate a higher revenue than the others,
on average?
As you can see the group who receive the newsletter has a average revenue of 59021 and
the group who do not receive the newsletter ownes a lower revenue of 51976. In the other
figure the p-value is ,000 and that is less than 0,05 (0,00<0,05). So we can conclude that the
consumers who receive a newsletter has on average a higher revenu. Thereby the null
hypothesis is not rejected.
2. Is there a difference in average age depending on the size of the house hold a customer is
part of?
By answering this question I used the ANOVA T-Test because the variable, size of the house
hold has multiple groups. P-value 0,000<0,05. So the first pictures says that the average age
on at least one group of house hold differs. But to confirm that I used the Duncan test and in
the second picture the p-value is on each group of house holds 1,000 and that is way to high
to confirm that there is a difference in average age depending on the size of the house hold.
As you can see the average age of the house hold size groups 1-2 persons do not differ
significantly. Also for the other house hold size groups the average age do no differ
significantly. So the null hypothesis is rejected.
2 Timo Rouwenhorst, 645803