Visualizing H-1B Visa Application

I worked as a designer in this team. The purpose of this project is to practice data programming skills and design skills in creating data visualizations for a selected group.

This is a class project from an IMT visualization design course at the University of Washington.

Duration: Mar, 2018 (4 weeks)
Role: Visualization designer; Usability tester
Team: 1 Researcher; 1 Designer
Skills: R; Adobe AI; A/B testing; Cognitive walk-through

Problem

International students need an H1B-sponsored job to enter the U.S. workforce. An employer must obtain a certified labor certification (aka. H-1B visa) application to hire a foreign worker to work permanently in the U.S. Since not every company is willing to offer sponsorship, students have to go through a series of tedious searches before applying.

Design Challenge

How to add and visualize the meaning and context of data?

 
 

Background

Data source: US Permanent Visa Application dataset

 
 

Audience and Motivation

I composed a persona to better represent our target audience. In a nutshell, the motivation, i.e. tasks to be completed using the visualizations, could be summarized as

“identify U.S. companies that offer abundant H1B-sponsored jobs which also meet the audience’s other expectations, such as salary, field of interest, and location.”

Breaking it down:

  1. What are the companies that offer the highest number of H1B sponsored jobs?

  2. To which economic sectors do these companies belong?

  3. Where are they?

  4. What is the salary expectation of these companies?

 
 

Ideation

Question 1, 2 & 4:

  • What are the companies that offer the highest number of H1B sponsored jobs?

  • To which economic sectors do these companies belong?

  • What is the salary expectation of these companies?

 
 

After trying various visual designs and encodings, we decided on a visualization integrating favorable features from previous ideations.

  • X-axis: average number of applications per year

  • Y-axis: average salary

  • Color: economic sector

  • Size: certification rate

 
 
 

Question 3:

  • Where are they? 

 
 

I combined the two color sets in the final design. Warm colors are assigned to states with higher H1B applications and cold colors for the lower ones.

The information that users can derive from the gradient color scale is vague and not quantifiable. Therefore, I broke down the color scale into color blocks and divided the data into 6 distinct buckets with numeric value annotated.

 
 

Usability Testing

After my teammate screened out the data, I used R to execute the drafts and refined them with Adobe Illustrator. 
Seven individuals were chosen as close representatives of our target audience, and below is the user profile I created. 

 
 

I also visualized the usability testing results. 

Final Visualization

Based on the user feedback, I modified and created the final visualizations.

Visualization 1

Reflection

This project spanned less than a month, which is a testament to how frantically rushed we all felt, but also shows how resourceful we are as a team.

As I learned a host of programming and design skills, the biggest takeaway from this project is actually the importance of documentation in design. Design documentation can communicate to people who may be interested in the project - what my design decisions are and why my decisions are worthy and reasonable - which is critical for further engagement. If time allows, I’d make the visualization dynamic, interactive, and easier to consume.