Case study
I worked with Carina, a care matching site linking licensed caregivers with people on Medicare or Medicaid requiring in-home care. My team and I evaluated Carina’s platform with a DEI lens to identify and address opportunities for bias or discrimination.
Our client Carina aimed to provide a more just and equitable experience on their care matching site. However they weren’t sure how to go about making that happen. My goal was to evaluate their platform from a DEI lens, and to identify and address existing opportunities for bias or discrimination.
UX design, UI design, User research, Project management.
Team consisting of myself and 3 other graduate students (1 designer and 2 researchers).
2021.
Research DEI best practices and the caregiving space.
Understand caregivers and their clients.
Update the profile experience to highlight caregivers’ humanity and overcome biases.
Deliver design recommendations and process to Carina, share learnings with the UX community.
My team and I began by seeking resources about DEI best practices, looking for a checklist we could use to evaluate Carina’s website as good or bad. We quickly realized such a tool didn’t exist and had to dig deeper. I read case studies and interviewed UX professionals about inclusive design. We then conducted stakeholder interviews and a competitive analysis to gain insight into the caregiving space.
Carina lacked awareness regarding the extent of discrimination on their site. They relied on a discrimination reporting feature that had only been used twice. With tens of thousands of users, we suspected the issue was much broader, but had no idea how deep it actually went. In order to investigate, I sent a survey out to Carina’s email list of providers.
The results were astonishing and disappointing. Over 35% of providers had experienced bias or discrimination on the job, with 13% facing it on the Carina platform. This starkly contrasted with the two reported incidents. While Carina had been focusing on reactive methods like their reporting feature to combat discrimination, they had missed the majority of incidents.
Shifting from reactive measures, my focus turned to prevention. Interviews with providers who had faced discrimination highlighted their reluctance to report issues due to job security concerns. My key takeaway was the need for a proactive approach to prevent discrimination. I needed to help Carina become more proactive so that their users wouldn’t be forced into a decision between keeping their job or staying with an abusive client.
In an interesting finding from interviews, we found that Carina’s previous attempts to reduce discrimination was actually limiting clients from finding good caregiving matches. Carina employed a technique called progressive disclosure, concealing demographic information like age and race from the profile to avoid bias during initial matches. For example, they hid profile pictures from search results. However, our research revealed the unintended consequences of this approach.
First, without insight into providers’ personalities and unique experience, clients struggled to overcome implicit biases based on the limited information that was in the profile. We found many clients making decisions about who to contact primarily based on gender. Second, hiding demographic details didn’t actually prevent discrimination—users were being shielded from discrimination online only to face it full blast when they met up in person or on the phone to interview. By failing to consider the user journey beyond the Carina platform, Carina had put their most marginalized users into dangerous situations.
When we talked to these users, they expressed a preference to present their whole selves on their profiles in order to connect with open-minded people and to allow bigoted folks to stay away. Based on these insights, my recommendation was for Carina to abandon their focus on progressive disclosure and instead give their users control over how they present themselves online.
When prototyping, my team and I focused on the caregiver profile. We wanted to demonstrate how to highlight caregivers as whole, qualified individuals in a way that helps users overcome implicit biases and find good matches. Our recommendations centered on redesigning the profile's structure and language, emphasizing person-first language and unique qualifications.
For example, in the previous design users could select from a short list of “caregiving superpowers” to highlight on their profile. Few providers filled this out because the options were too limited to express their unique skills. Our proposed design used the care provider’s name to refer to them throughout and restructured the profile form to be less restrictive. We also renamed “caregiving superpowers” to “unique care skills” and expanded the capabilities of the skills section, emphasizing what care providers bring to the table rather than what they lack.
Final design links (password protected):
Carina integrated our design suggestions into their development roadmap and initiated the DEI-focused design process modeled by my team.
This project truly demonstrates how I approach design and what I value as a designer. I care about accessible and inclusive design, which were both a major part of the project. Furthermore, I took a forward thinking approach by not treating this as a standalone project but took the opportunity to share my learnings with this Interaction DEI Toolkit. My aim with this toolkit was to empower others to follow a similar process for creating equitable and inclusive designs. I have already adapted this toolkit into my roles at different organizations.
