The Analysis
So the attributes in this set are the card type (Best Buy, iTunes etc), the value of the card, how much it sold for, what were the shipping costs, how many bids did it have, what was the feedback rating of the seller, the percentage of the sale (including the shipping), the average percentage per card and the actual link of the auction. So that means there are a large amount of things you can analyse. For single variable stuff you could find measures of central tendency for the entire set or individually for each type of card. Or just choose your type of single variable graph and create it for the whole group or by card type.Or you could do some double variable analysis comparing to see the connection between the value of the card and the sale price (for either the whole group or by card type.
And because the data exists, you could even do some comparisons of the average percentage that a card gets.
Sample Questions
- Identify the outliers for each card type (Value, sold etc) and suggest why they might be outliers
- Identify the spread for the Value of each card type. Why might some cards have smaller spreads than others?
- How does the linear regression compare for different types of cards?
- Are there any cards that were sold for more than they were worth? What might cause someone to pay more for a card than what it is worth?
- Why might some cards have a higher average sale rate?
Other Stories
This data came out of a story originally about why companies love gift cards (and the page of supporting data for the article) As it turns out they actually tend to be like free money. This is because so often people don't use up all of their gift cards and then forget about them. I think part of that is because we are required to know exactly how much is left on a gift card in order to use it. They actually show the data (on pg 65) for Best Buy on how much extra money they made because of unused gift cards (spoiler alert, it was $43 million)Download the Data
- The original Spreadsheet (Excel, Google Sheets)
- Fathom (Data, With Graphs)
- CODAP file
No comments:
Post a Comment