Visualizing Uncertainty- - - - - - - - - - - - -
I prototyped a minimally interactive data visualization that concretely displays uncertainty for a non-expert adult audience. The images serve as one screen amongst many to be viewed in a website format.
Data for the project was derived from a fivethirtyeight poll about how popular or (un) popular president Donald Trump is over the course of his presidency. Polls can be viewed in group format, solo, or in comparison with another. Below are some examples of original poll data:
For my design, I decided to use the color saturation and negative space to show areas of uncertainty. The more saturated the color, the more heavily weighted the polls are the less saturation, the less reputable they are and thus, less weight they have in influencing the data. The areas of overlap show areas of some certainty and areas where all circles overlap show the most certainty. In this case, dark, heavily saturated space indicates a "most certain" area of agreement among pollsters.
By allowing the user to shift through two-poll comparisons and view areas of certainty and uncertainty side by side, they are able to see a drastic visual difference accounted by the negative space. The sudden removal of visual material can evoke a more shocking, emotional reaction to the amount of certainty among pollsters—helpful for the layman who might not want to spend a lot of time trying to read between the lines and decipher complex poll information. The goal is to show them uncertainty quickly through simple visuals and through the power of negative space.
In stressful times such as election season when there is a lot of data moving around trying to "predict" outcomes, it is important to show all sides of the story and to provide clear, evocative visuals to the user who is already oversaturated with information.
