I used Microsoft Office Excel to create a chart, specifically a histogram, to show the discount percentage of a surprise sale of Kate Spade New York.
Because of some technical issues with my computer, I can't upload pictures to Gallery.
All the pictures I intended to use is in this file: https://drive.google.com/a/andrew.cmu.edu/folderview?id=0B0sNjjqHXHnxWWdSLTByWTVRV2s&usp=sharing
When I see the title "lying with data", I immediately think of the sales advertisements companies make to attract customers. Usually they lie with the discount of products to make customers feel like they just have to buy something because the discount is huge. However, customers now are aware of the lies of sales. Most of them know that a "up to 70% sale" does not necessarily mean every item they but is 70% off. Therefore I want to show my data in a way that convince customers that their is in fact, a 70% off and sometimes even more than 70%.
I received an email from Kate Spade last week saying that there's a surprise sale online. When I went to their website and saw the "up to 70% off" slogan, my instant thought was that it must be a lie made by the company to persuade me to buy more and spend more. This made me want to gather the data and find out whether "70% off"is a lie or not.
There are some analysis on the actual discount in sales, but I didn't find one similar to my project.
Since the official site of Kate Spade does not provide the data of discount percentage for obvious reasons, I had to type in every price myself. I used Excel to create a form and then type in the original price and discounted price of every object. Fortunately there were only sightly over 100 objects so it didn't take me too long. Then I created a histogram to show the discount percentage. There were several forms of histogram, I tried them all and picked the neatest one, which I think can help viewers focus more on the data itself rather than the how it is represented (Just like Tufte mentioned in his article).
The data I typed in was in random order, at first I tried to range them in descending order, but then I realized that makes the histogram too straightforward. By creating a randomized histogram, I feel like viewers are more likely to focus on the data that stands out (like those items that is actually more than 70% off). In this way, viewers will be more convince that the sale is actually 70% off.
This is the hardest project for me. At the beginning, I'm not sure how should I lie with data, since data itself is always true. But after reading the links provided in the google doc, I got some ideas and eventually created this project. I learned that different ways to represent the same data may results vastly different effects.
You can upload files of up to 20MB using this form.