Another variable was the delivery charge. Some supermarkets didn’t have one, and others had different charges depending on how far away the customer lived. I averaged out the delivery charge, where necessary, for each supermarket.
Primary/Secondary Sources of Data
I decided to use primary data, as I then knew that an un-biased person had collected it, and I knew that my data was accurate. The only drawback was that it took me quite some time to gather the data.
If I had decided to use secondary data, it would most likely have been much less time consuming and I would probably have had more data. There would also have been the problem that the data might not have been accurate (as the person gathering it may have been careless or biased) and it might have been hard to gather the exact data that I required.
Methods of Collection/Selection
I searched for, and found, four different supermarkets that allow their customers to shop online and listed the prices of my chosen 30 items.
I collected my data by searching on the Internet for supermarkets that offered online shopping and delivered to homes in England.
I had to use a sample of standard groceries that would be representative of a typical weekly ‘basket’ and that each supermarket would sell.
Sorting Data (Criteria)
The data were sorted in terms of the type of grocery, the cheapest and dearest cost and the identity of the supermarket.
Control of Variables
The variables were the prices of different brands of each item, and the delivery charges.
These were controlled by choosing only the cheapest and most expensive brands for each item and by averaging out any variance in the cost of delivery.
The cost of delivery was in some cases dependent on the location of the customer relative to the store and also to the time slot chosen for delivery.
Data Tables
These are the cheapest prices of each item:
These are the most expensive prices of each item:
Graphical Methods & Interpretation
There were 120 items of data. In order to make some sense of the numbers it was necessary to order them in various ways.
Clearly a simple addition of the costs for each supermarket, and for their respective cheap and dear baskets would allow some graphical representation.
Ordered bar charts were chosen as the method most quickly and easily understood by the reader.
Thus one can see that when the raw data are added to give total costs for the baskets, there is a different ranking for the subject supermarkets. Iceland is the dearest for lower priced goods, even without a delivery charge. After saying that, for the premium goods it is the cheapest. I had noticed in gathering the data that Iceland sells fewer ‘own brands’ in comparison with Tesco, for example. So this might be an explanation for this paradox.
When the prices of the cheapest and dearest products are averaged out it can be seen for example that the position of Sainsbury moves from second ranking to the dearest for the full ‘basket’.
It became clear then that simple comparisons such as those above would not be sufficient to fully inform the consumer. More detailed analysis was required.
To do so I needed a graphical method to show the spread of the respective groupings of data in terms of ‘cheapest’, ‘dearest’ and average shopping baskets. See Annex A.
The ‘box and whisker’ diagrams show that there were larger ranges and inter-quartile ranges for the supermarkets with respect to the totals for the dearer shopping basket. The average and cheap baskets were more narrowly ranged.
This may indicate more vigorous price competition for goods at the ‘budget’ end of the spectrum.
Analysis
The next question the consumer could be asking, following on from the interpretation above, might be whether the cost of the baskets was skewed one way or another by particular ‘rogue’ items. For example, if one stopped buying baked beans from a particular grocer, if one was to be more selective in the items chosen for one’s ‘basket’, might this have a significant effect on the relative costs of the respective baskets.
More analysis was required in an attempt to answer this question and address the hypothesis in a more intelligent manner.
I decided, therefore, to use the measure of deviation from the mean for each grocer and each commodity to investigate whether dispensing with particular items from the basket in each case would affect the decision as to whether one supermarket is cheaper than another.
This graph shows that the delivery charge greatly affects the total price where Iceland is concerned as it shows that the majority of Iceland’s prices are above the mean price. This is hidden by the fact that they are the only shop that doesn’t charge a delivery price.
The graph also shows that the price of washing powder varies greatly from shop to shop. It is on the outskirts on all four graphs, showing that the range of prices is somewhat varied.
Milk is one of the goods that is always close to the mean, showing that it is a product whose price doesn’t vary as much as other products, for example, shampoo. Shampoo is always on the outskirts of the graphs, so the graph is showing us that, like washing powder, it is a product with prices that are varied.
Conclusion
“Of all the supermarkets that offer shopping online, it is more expensive to buy an average ‘basket’ of groceries from Sainsbury.”
The price of a cheap ‘basket’ at Sainsbury was one of the lowest, and so was the price of the dearest ‘basket’, but when one looks at the average prices, Sainsbury comes up as one of the most expensive (second to Iceland).
When deciding whether I have proved or disproved my hypothesis, one has to take into account what type of shopper one is. A budget buyer, buying all the cheapest options, or a premium buyer, buying the expensive options, might do well shopping online at Sainsbury, whereas someone who tends to be a mixture of the two, would do better shopping elsewhere.
The delivery charge is greatly significant as it masks the fact that Iceland’s prices were rather costly compared with others. That is the price of ordering food from the Internet though. If one were to compare the prices of groceries bought straight from the shop, we would see a big difference as the delivery price is, in one case, around 18% of the overall price.
I was fortunate, as I didn’t encounter many problems during my investigation.
I do, however, think it would have been better if there were a greater variation of prices. Some of the prices were nearly the same and I think it would have been better if there were more of a variation for me to comment on.