Missie 9: Outside looking in
Recently, a group of students were talking about AI misgendering black women and being unable to identify Asian people. They remarked: “Maybe these functions weren’t intended for them, so they shouldn’t complain.” To say that I was shocked is an understatement. How could they think that?
Six months before, another group of students and I discussed the term ‘Data Bias’, and the effects of it in our society. I was surprised to learn that most of them did not seem to care much, as they didn’t identify with being excluded. I asked the students to create user profiles and all their intended and expected users were “white”, ‘male’, ‘able-bodied’, ‘educated’ and ‘employed’, a complete mismatch with the students in my classroom.
Why was this happening? I learned that most interpretations of diversity and inclusivity don’t seem to cover the true breadth and depth of these terms, which poses a problem when designing software that’s meant to be inclusive. One has to wonder, how can you be inclusive when you don’t know who you are trying to include?
At HEMD, one of the things I focus on is designing conversational tools for sensitive topics such as mental health issues, sexual abuse, maternal and child healthcare. These are systems that are very sensitive to their users’ needs, which may vary based on age, race, gender, cultural backgrounds, religious beliefs, sexual orientation, physical and mental ability, socio-economic status and literacy level. And the list goes on.
You’re right, that’s a lot of parameters to consider. And yes, designing inclusive systems can be more difficult than designing systems that simply don’t take these things into account. But the consequences of designing systems that don’t connect with their audience – or rather, failing to design systems that do – can be very painful.
For instance, many services vital for life in the Netherlands – government websites, health insurance, online banking – lack other language options. Leaving me lost in my work and daily life. Some days I think if it weren’t for Google Translate, I would’ve signed away my right to live. I don’t understand half the information I’m supposed to, because despite up to 2.5 million people not speaking Dutch ‘adequately’, very little is being done to facilitate them.
Thus, we come to the topic of the mission we’re introducing today. The question of ‘how can you minimise exclusion through design?’
The words used there were chosen very carefully. Because there are, in fact, legitimate reasons to design things with a specific target audience in mind. We’ll get into that topic another time. The point is: Whatever you design should not exclude users interested in using your product.
We want to find out what prevents users from connecting with both digital and physical spaces and products, and find out what can be done to improve their design to be more in line to the needs of its users. We want to look at who is doing it right and where mistakes are being made, as well as how to prevent them.
I’m undertaking this mission and the research backing it with Marissa Berk, one of my colleagues at HEMD who you might have met previously in the presentation about the GDPR tool that HEMD developed. She also has a background in user-centered and data-driven design, and we’ll be working together closely throughout the process.
One of the major challenges for the project will be to find a group that is diverse enough to provide us with the variety of perspectives that we want to consider. It’s easy to claim to be inclusive, but to actually test your designs with a user base that is not just diverse in gender, ethnicity, and religion but also in educational background, income level, mental health, age, sexual orientation, physical and mental capabilities, linguistic skills (and more), can be extremely difficult. The list of different types of people is never-ending, and so identifying what characteristics influence which aspects of user experience will be an important piece of the puzzle.
There’s a lot of things that aren’t yet set in stone, and our next steps will be to talk to other researchers and designers working on this topic, to see what things they might be struggling with.
We hope you will join us in the coming months as we explore this topic further. Follow us on Instagram, LinkedIn and Youtube to stay up to date with our latest releases, and of course, don’t forget to keep an eye on our website either.
De producties voor deze missie worden ondersteund door redacteur Sander van Velze.