Caffeine Data Visualization
Here is a good example of someone trying to get creative with displaying the amount of caffeine in different types of drinks. I applaud their effort because it’s different and visually ties the amount of caffeine in each drink to a brand.
Below is the original data visualization:
One issue I see is in looking at Pepsi Max and Sobé, which have 46 and 48 mg of caffeine per serving respectively. The problem lies in the bar displayed in Pepsi Max that shows the 46 mg being higher than the 48 mg bar. I can only guess how that may be possible, but think it may be an error or oversight.
Below is my rendition of a better design:
I have no way of getting rid of the bars they used in the drinks. If possible, I would have removed the bars and added them to below the picture as shown in my version above. Also, I removed the names of the drinks because they are self-evident. Granted, this data could be shown in a simple bar chart, but adding the pictures makes it a little more eclectic.
Most Commented Posts
December 3rd, 2007 at 5:14 pm
It’s inherently a little bit misleading to use photographic representations of objects of different sizes in a context like this. There’s a strong natural inclination to assume that the images that appear larger are designed that way to indicate higher caffeine content. So at first glance, or certainly in the thumbnail versions of the images, the viewer is inclined to think that the Propel, Coca-Cola Blak, and Pepsi Max products are the ones with the highest caffeine contents. You have to look a little more carefully to realize that the photos are just photos, without any data content, and that the only actual data content is contained in the bars superimposed on the photos.
December 3rd, 2007 at 9:35 pm
Jonathan – I definitely agree with you on this one! I have seen many uses of images/graphics to depict data and often they fail miserably (some examples are: houses to show new home starts or sales, cars to show car sales, etc.).
One of the main reasons they fail is because they are using area or volume to show one-dimension data. Edward Tufte devotes part of chapter two in The Visual Display of Quantitative Information to the idea of graphical integrity, which touches upon this subject.