1000 word individual report and associated visual outputs (1,000 word
equivalent) based on practical work, due before 2pm on Thursday 22nd
February (Week 18); 50% of module mark.
This assignment is based on the Dunnhumby data called Carbo-Loading:
A Relational Database as described
here https://www.dunnhumby.com/sourcefiles .
You will need to download the data provided here (you could do it from the
site). Please contact Lex if you have any problems doing this!. The data
includes 4 datasets in CSV / SAV formats and a User Guide. You will need to
familairise yourself with the commands for linking datasets described in earlier
practicals.
The aim of this project is to answer 3 questions, of increasingly complexity:
Q1: What is the household penetration of ‘PRIVATE LABEL THIN
SPAGHETTI’? That is, out of all customers purchasing Pasta, what percent
purchase this brand? (10 marks)
Q2: How does the household penetration of the product PRIVATE LABEL
THIN SPAGHETTI vary within the two regions, relative to the sales of Pasta in
each region? (20 marks)
Q3: For which category of products (pasta, pasta sauces, syrups and
pancake mixes) does the provision of coupons appear to have a positive
impact? For this you should consider which customers first purchased an item
in a category using a coupon, and then how many of these customers made
additional purchases of the item in the category. (30 marks)
Up to 20 marks will be awarded for the descriptions of what you did
(remember figures and screen grabs of code are not included in the word
count).
A further 20 marks will be awarded for presentational clarity,
visualisations etc.