Recognizing Diversity and Bias in Research and Design

Author Date June 26, 2019 Read 8 min


projekt202 UX Researcher Marta Zarzycka and Experience Designer Manimala Karusala discuss diversity and bias in research and design.


Note: To watch with closed captioning, hover your mouse over the video above and click the CC icon. You can also follow the transcript below, whichever option is more convenient for you.


Hi, my name is Marta, and I am a user experience researcher here at projekt202.


Hi, I’m Mina, and I am a UX designer at projekt202.


And today we will talk to you about the diversity and bias in research and design.

Bias in research and design has been perpetuated by Silicon Valley almost every day, and only now we are becoming increasingly aware of it. It can range from the lack of empathy for certain groups to major cases of social injustice.

And the examples are alarmingly easy to find. You can think for example of AI technologies that inform facial recognition that are perpetrated by social media for ads and are targeting or calibrated based on people’s gender or race or socioeconomic status.

Another example is a recent study carried by an MIT researcher that shows that face recognition systems by Microsoft or IBM are far less accurate when used on women or people of color.

Another example is for instance dating apps. The majority of dating apps are designed for people that are white, heterosexual, cisgender, and able-bodied. As a result, people that are anything but lack safe spaces in those apps.


Yes. And bias often affects our everyday lives as designers because we end up designing systems that seem to do everything right, but often end up providing a bad experience.

Take screen readers for example. People with vision impairment rely on screen readers to read the content out loud for them. The way these screen leaders interpret emojis is that they describe every emoji as it appears.

So imagine how popular emojis are right now and how frustrating it must be for someone with vision impairment and their screen reader has to read a text message that only consists of emojis, or a message followed by 10 smiley faces.

The takeaway here is that bias doesn’t really stop with broad topics like gender diversity or race. It really prompts us to look deeply into how people’s way of life affects the way they interact with their environment and the decisions they make every day.


How to unbias our own practice is a question we should be asking ourself almost every day.

For me personally as a researcher, it would be positioning myself in my research, but at the same time moving away from myself in my research.

So I constantly have to be aware to what extent my gender, my race, my age, my socio-cultural status, my socioeconomic status influences what I do and influences the conclusions that I reach, how do they influence my standpoint and to what extent they actually form a bias.

I also have to think and consider what kind of resources I have available and not necessarily take them for granted.

For instance, I have cutting edge technology. I have financial resources to do the research. I have a college degree. I have transportation capacity. And finally I have able body. I have to be aware that a lot of my users do not have the resources that I have in order to carry this research and constantly correct the assumptions that I am making throughout this process.

Moreover, we cannot always choose our clients. But if it’s possible, I always try to aim for a diverse representation of a client body and diverse representation of the stakeholders that I get in touch with.

In the same note, I always try to think about diversity when recruiting my users. There is no user experience without the body, and to ensure different kind of experiences, we also have to take into account different kind of bodies. So when recruiting, I always think about the gender, age, economic status, tenure, and body ability, race and ethnicity of my participants.

Thinking about my users, I try to think about the person with most needs first and then design for others will also benefit. When I think about the user journey, when I think about the customer journey, I never try to whitewash it. I always try to think about possible glitches, hardships, disadvantages that those users might have.

So I never imagine a customer journey as something that is smooth, easy going, and automatically getting rid of any kind of obstacles. Not every user journey is a happy journey. So I always try to think about less happy paths and drawbacks when I design for a client.


Yes, exactly. Even as a designer, it’s not easy to completely unbias ourselves, especially because we’re often running up against deadlines, we’re running up against market trends and release dates. But if you start small, it definitely can be done.

We can start by making sure that we’re identifying obvious points of exclusions such as color blindness and deafness and try to provide solutions such as high-contrast text or closed captions. These accessibility guidelines are very easy to implement and the user can actually set these themselves to suit their own needs.

We should also understand that our products will probably not be used in a sterile environment. So people might be opening up our app on low-end devices. They may be in low light, noisy areas. They may not speak your primary language, or they may speak it with a different accent. But nevertheless, our experiences still need to make sense in their context.

Ideally we should reach out to diverse groups and get out of our personal comfort zone to gather firsthand feedback. But if that’s not always possible, you can simulate it yourself.

For example, if you’re developing a voice messaging app, try to see if still works if you’re in a crowded coffee shop and people are moving around and there’s a lot of noise coming in.

And finally, let’s pay attention to language and visuals. Using gender neutral pronouns and respectful ways of addressing your user can go a long way toward promoting inclusivity and goodwill from your customer.

And making sure that your user base is represented in promotional photos and advertisements and splash screens.

Also check action photography. In these action photos, who is always center stage, who always looks like they’re the most knowledgeable, who always is given the center shot while others seem to be cropped or marginalized? So making sure that you’re aware of all these things can really put you on a good path to unbiasing yourself throughout your career.


The more we recognize how bias affects our research and our design, the better positioned we are to account for it and also to minimize its impact. One way to counter it it’s collaboration. First and foremost, in this case, collaboration between researchers and designers.

It’s important to go out of your way to be exposed to people with different views, with different backgrounds, and different beliefs. It’s important to surround yourself with people that promote diversity. It’s important to attend conferences, meetups, and networking events that strive for equal opportunities and also fight for disempowered communities.

Also participating in in-house shadowing and outreach programs is something that can help you expand your horizons and make you aware of your own bias.

I think you should also address diversity not only among clients and users, but also within your own team. If your team is too homogenous, if it’s too much like, if people resemble each other too much, you are likely to receive responses that only confirm your bias. So in a way, seeking out the disagreement within a team is something that is very important. The greatest danger of any bias is that it discounts and ignores contrary views.


This is actually a really great question because as user experience designers, we really need to realize that we are in charge of creating one of the most powerful and influential deliverables in the organization, which is this product experience. And as such, we have access to the Dev team. We have access to the stakeholders. We have access to decision makers. We also have access to our users.

So we should use this access as a way to promote best practices. Accessibility and inclusivity should be a baked part of our process rather than a special case that requires extra attention.

So first things first, learn to speak up and address when there are gaps in requirements that ignore edge cases. Sometimes stakeholders will try to tell you, “Oh, that’s only 8% of the population. Does it really matter?” If you have 10 million users, that’s 800,000 people. So yes, it really does matter that we take edge cases into account and make sure that the experience works for as many people as possible.

Not to mention that these accessibility points, they are very cheap to implement, but the goodwill and the loyalty that they generate, it’s almost priceless.

As designers, we should stay ahead of how addressing and eliminating bias can actually make a better product and generate more revenue. If we frame arguments in this manner, it will be very easy to get buy-in from stakeholders, and the next time it will be much easier for you to make arguments saying that we should be more accessible, saying that we should be more inclusive.

Once again, the goodwill and loyalty and recommendations that we garner from our users if we continue to serve them is almost priceless.

And finally, try your best to publish your learning moments. Whenever you come across a use case that you’ve never considered before, find a way to put it online and share it with the community. This is great not only for others to learn from you, but also to give you insight into any problems you may be facing.

Putting this online is also a great way to collaborate with people who you may not have immediate access to. So designers who are working in that same problem space may be willing to pitch in, or users who have firsthand experience into the problem that you are experiencing may be able to pitch in as well.

All of this really builds a community. And the more data we have, the better we can make the technology. As we gather more data from more diverse sources, we’ll have a more complete understanding of what problems people are facing and be better equipped to solve them.


We hope that we have encouraged you today to think critically about bias that you might face in your own practice, and also to make you think how you can counter it.

Find Your Possible.

Let's Chat