A Designer's Landscape
(under construction) An interactive experience that allows users to compare their visions to the 2017 AIGA Design Census data.
The Design Census is an open and collaborative survey conducted by AIGA in 2017. There are over 13,000 responses collected in the census. Using the design census data, we wanted to answer these questions:
What is the state of design today?
What are the methods to analyze and visualize data?
How can designers use data beyond the purpose of information delivery?
To get the full interactive experience, please visit the website here.
For the best experience, please view the website with 1440x900 window size.
How it works
Raw Data Analysis and Ideation
How could we find meaning in 13,000 entries of raw data?
Our raw data mining process intermingled with our ideation process, because our concept would fundamentally affect the ways we could analyze this huge amount of data. After some initial grouping of the categories of data available from the census, we have came up with several different ideas that could make good usage of the data:
With the initial questions in mind, we decided to compare the user input with the data to allow user the data beyond the purpose of information delivery. Taking the current types of responses we have, we categorizes them into “qualities that measure success“ and “qualities that can lead to success“. Qualities that measure success include: salary, year in design, job stability, and job satisfaction, and qualities that cause success includes skills I am good at, side jobs, skills that I want to learn, and current field of work. The graph above shows some initial analysis we did to compare the differences between all designers and the successful group. view our initial pitch slideshow
But… who define success?
At first, we struggled for a narrow definition of “success“. We realized that a singular measure of success may cause a lot of disagreement for different users. But what if the users could define their own successes? To keep the number of types down, we limited the choice to 2 out of 4 options (salary, longevity, satisfaction, stability), which yields 6 types of success. Then we re-assessed the data for the four questions that we nailed down, as shown above. It was interesting to see how different types of successes yields such different results.
Do we want the representation to be more informational or poetic? Where does our visualization lie on this axis?
We attempted to find out how much information a visual piece should carry in our visual explorations. Around the idea of map, we tried different types of assets. Our first iteration was all animals, because we thought the idea of food chain would explain which choices would lead to success and which would not. But since it would be a little too obvious for implying “survival“, we decided to move to objects like buildings and trees. Each set of asset has a hierarchy, and is matched to the data that is compared with the overall data. The placement of assets strictly follows the stats, which was made into a reference chart below:
User Experience Design
For our screen iterations, we explored different user flows that would make the interactive experience more consistent. One major discussion was around whether we should incorporate the form with the map. We decided to have a fixed position for the form, and an automatic transition from section to section in the map.
How to increase the visibility of data in this metaphoric experience?
One challenge we encountered while designing this experience was that the numbers in the data that controls the algorithm is not transparent. Therefore, we added in the interactive hovers that would illustrate what each part of the map represents better.
User Testing and Feedback
Moving forward, we are looking to improve:
Raise awareness of the contrast in result between different success types
A motion piece that explains the algorithm thoroughly
More interactivity during the transitions between the questions
About Data Visualization
I learned that to visualize data in a meaningful way, first we have to find meaning in data by looking at the data through different lenses, and then we need to find a balance between representation and information. I think there is no definite right or wrong for where the visualization should lie in this axis, but there are always better places on this axis to support a specific concept.
About The Process
To develop an experience from frontend to backend was not an easy process. I think that our design process in this project was a lot different from my past product design experience because we had more constraints from the backend. Although we first proposed to settle down the visuals before we began developing the site, the process ended up being more sinuous and we had to keep coming back to refine the visuals to meet our new backend challenges. It was also interesting to be in a team with two other designers, and we each had our strength but also some overlaps. We were able to split the work but also pass around responsibilities to create a more consistent experience.