Creating a Data Visualization Website for Farmworker Advocacy
Team: Cornell Hack4Impact — Designers: Helen Li, Anna Nguyen,
Project Manager: Miyuki, Tech Lead: Kevin, Developers: Rahul, Ananya, Kenny, CJ, Hannah, Amy
Timeline: Fall 2021 (10 weeks overall)
Farmworker Justice is a nonprofit organization that seeks to empower migrant and seasonal farmworkers to improve their living and working conditions, immigration status, health, occupational safety, and access to justice. Farmworker Justice works with farmworkers and their organizations throughout the nation.
The Problem
The National Agricultural Workers Survey (NAWS) is an employment-based, random-sample survey of U.S. crop workers that collects demographic, employment, and health data in face-to-face interviews. Currently, many people use NAWS to advocate for farmworker rights, including community partners (legal services, advocates, organizations), government officials, researchers and academic institutions, journalists, and administrations with specific research questions related to farmworkers. However, NAWS data is only available in a codebook format, with no data visualizations offered. There is a data summary report based on the codebook containing visualizations; however this report is limited in usability and is not aggregated every year, proving unreliable. As a result, many people reach out to Farmworker Justice for interpretations of the data, which can be time consuming.
The Solution
We are creating a dynamic data visualization platform for Farmworker Justice and allies to easily access NAWS data in a more readable manner.
User Research
We wanted to better understand the types of people accessing NAWS data and for what purposes. Additionally, we wanted to understand people’s process of accessing the data and interpreting it for their own means.
We also asked users to explore an existing data visualization page with information about Cornell University’s Doctoral Program Statistics in order to learn more about the user’s mental model and the way users visualize multiple pieces of information.
Key Insights
- The data is not substantive enough and does not dive deep enough into state-level arguments, or specific populations
- Users looked at the summary report first to try to find what they need
- Users found the broad categories helpful as a high level overview, but need to go deeper into subcategories to find what they need
- Users have different comfort levels working with the raw data depending on their background knowledge.
Information Architecture
Minipage
Key Functionalities
A minipage hosts all data visualizations in a given category of NAWS survey data.
These data visualizations are generated with data collected on different NAWS survey questions.
Users can search for data visualizations of their interest by the survey questions or keywords in the survey questions.
Users can user filters to determine which data points are included when building the data visualizations.
Design Decisions
How much information should we display on one screen?
The first design decision we had to make was to determine the number of graphs we display on each screen.
While showing multiple graphs could provide an overview of the category at a glance, showing one graph at a time would help users better focus their attention on the details.
After several rounds of user testing and discussion with our partners we decided to go with single graphs for three main reasons:
- Users report being distracted when there were multiple graphs on a screen.
- Single graph means bigger graphs, which improve the accessibility of our site for senior users.
- Single graph may work better with our filter feature as users were often confused if the filters were applied when we user-tested with multiple graphs.
How should filters work?
Next, we explored different designs of the filters feature. The main goals of these explorations were to figure out:
- How might we clearly indicate which graphs are filtered?
- How should the filter panel look like?
- Should filters be applied to all graphs or just selected graphs?
After comparing and contrasting many versions and getting feedback from our partners, we eventually decided to:
Use filter bar instead of panel because:
- Minimally interfere with viewing the graphs
- Shorter list of choices would appear less intimidating for users who are less familiar with digital interfaces
Filter all graphs at once instead of allowing users to filter selected graphs because:
- Less decision-making required before users could apply a filter
- Similarly, simpler interactions would also make our interface look less intimidating
Final Minipage Design (Mid-Fidelity)
Conclusions
It was truly a pleasure to work with our partners — Alexis and Andrew — from Farmworker Justice. The work that they are doing to help journalists and advocates improve the working and living conditions of farmworkers is remarkable and we were honored to become a small part of it.
Besides contributing to a great mission, we grew a lot as designers ourselves as we overcame hurdles and made complex decisions. For instance, we experimented with using other data visualization websites for user testing to garner insights for our own project. We also learned about how to design simpler and more accessible interface for users who are not familiar with digital interactions.
As our project continues to the Spring 2022 semester, we are hoping to test and improve our designs to ensure maximum usability for our future users.
Thank you for your interest in our project! :)
