When learning and analyzing data, it is crucial to recognize the human bias and error that can intentionally (or unintentionally) manipulate the viewer.

I gathered data from a peer tracking their stress levels, caffeine intake, and hours of sleep per night. I decided beforehand to make the argument that more caffeine and less sleep makes you less stressed, and designed the result with this argument in mind. In addition to the poster, augmented reality was used to reveal the actual bias behind the argument I was making, and the data I excluded on purpose.

You’ve been told caffeine is “bad” for you. You’ve been told around 8 hours of sleep per night is the “perfect amount”. But we all know, that combination is incompatible with nearly everyone’s lifestyle. Thankfully, that myth of “good sleep” and “natural energy” has been debunked!


What happens when every day of data is included in the first data visualization:


What happens when the ratio on the third data visualization changes its 1:2:2 ratio to 1:1:1: