Design for Understanding: UFO Sightings in the US

5 min readMar 2, 2023

Design for Understanding: UFO Sightings in the US

https://www.peakpx.com/en/hd-wallpaper-desktop-gktxp

Introduction

In this project, my team created two data portraits for a CORGIS dataset that recorded >80,000 UFO “sightings” around the world. The two portraits are an analysis view and a persuasion view. We want to use the most intuitive and clear format to demonstrate the three most important dimensions of UFO “sightings” — location, time, and description. For analysis, we employed Tableau Dashboard to showcase the geographical and chronological distribution of all UFO sightings while also including a word cloud telling the UFO shapes. As for the persuasion, we portrayed the number of UFO sightings in the form of a point cloud, showing an increasing number of sightings as time goes on.

Dataset

The dataset we used for this project is selected on CORGIS and it is originally from NUFORC, cleaned and uploaded to Github by Sigmond Axel. The cleaned dataset mainly consists of sighted date, location, duration, shape (description), and documented date, with over 80,000 entries.

Design Process

Analysis (Part A):

Our analysis design process mainly followed the Five Design Sheet process.

Brainstorm is the first step. We outlined all the possible visualization types for UFO sightings. Six chart types were brought up during the brainstorm, including dot map, choropleth map, bubble map, area graph, donut graph, and nightingale chart.

After brainstorming, we drafted the initial designs in order to perceive the options more straightforwardly. The graphs and charts come from the previous brainstorm process, but we generated more ideas during this step, for example, the line chart and the bar chart.

In the last step of FDS design process, we decided to use Tableau as the analysis tool and realized the final views based on the Tableau functionalities. The dot map with two filters (year and state) here presents individual sightings distributed across the nation with color illuminance encoding the time of the sightings.

The choropleth map displays the “density” of sightings in each state. The more sightings the darker color.

The line chart is perfect for showing the trend of the number of sightings throughout the history of records.

The bar chart leverages different color legends to show the different shapes of UFOs witnessed in each year. However, we did not include it in the final version, because the bar chart is not as straightforward and impacting as the line chart.

The finalized overall view is drafted here.

Finally, we implemented the visualizations using Tableau Dashboard. Here are a few advantages of Tableau: most efficient map implementation, integrated low-code platform, coordinated worksheets, and refined details (including tooltips and filters).

Check out the analysis visualization here.

Persuasion (Part B):

For persuasion, we first came up with the idea to persuade the audience with the story that UFO sightings are increasing every year in the US and that the existence of aliens is possible.

Based on what we created in the analysis part, we initially explored possibilities of using more charts in an intuitive order to convince the audience. However, we later realized that there is a more efficient and natural way of perceiving data — by seeing them directly. We decided that point (dot) cloud is the best way to most intuitively display UFO sightings, with each dot representing a sighting recorded in history. Point clouds can facilitate top-down processing, which is the process by which our brains use prior knowledge and expectations to interpret visual information. Because point clouds provide a high level of detail and accurate spatial information, they can help us recognize familiar objects and patterns more easily, including geographic and quantitative dimensions. Because of the way human brains process visual information, point clouds can be transformed into interactive experiences that provide a high level of realism and interactivity, making them much more immersive than charts and graphs. See Guided Search Theory.

Point cloud using quantity to convey the corresponding number of sightings in that year.

Another idea is to add the US map as background and translate latitude and longitude to canvas coordinates. This version is aborted for its complexity in the visual information.

The pie chart was discussed but discarded as well.

The Choropleth map is similar to the previous Tableau implementation.

Finalized view of our point cloud.

Finally, we implemented the point cloud using Javascript. As time goes from 1930 to 2014, each year’s UFO sightings are displayed on a dark canvas with the same exact number of dots. The colorful dots float around the screen as if they were real UFOs being witnessed by people around the US. We also added a background sound effect (spooky alien music) to convey the idea that the dots are symbolized UFOs.