Building with Purpose: How to craft a PhD in ethical tech.

August 16, 2022

Emily Wang '16 and Sara Hendren, Associate Professor, Design, Fine Arts, and Humanities at Olin, have written the following Q&A about accessibility and technology, finding the right graduate program, constructing a dissertation in disability and tech, and much more.

Pictured (L to R): Emily Wang '16 and Professor Sara Hendren. 

1) First, hi Emily! Tell folks a little bit about yourself: Where you live, what you’ve been up to since Olin, what your broad fields are and where you’re headed next.

Hi Sara! Thanks so much for inviting me to chat, as well as mentoring me throughout both undergrad and the recently completed PhD! I’ve been living in Chicagoland for the past 5 years, working on my PhD at Northwestern University. My doctorate degree is in Technology and Social Behavior, which is an interdisciplinary program in Computer Science and Communication Studies. Before that I was an engineering major with a concentration in computing (E:C) at Olin College. My work at the intersection of Human-Computer Interaction and Disability Studies focuses on how to design technologies and workplaces to be more accessible to people with disabilities.

As a member and collaborator of different disabled communities, I engage with disabled perspectives early and often throughout my research process. Every project I've done has been with a different disabled community and technology—I’m grateful to have learned together with Blind, Deaf, chronically ill, and neurodivergent collaborators throughout my career. This summer I will be starting as an Assistant Professor of Computer Science at Oberlin College, where I’ll be teaching classes about computing and inclusive design as well as continuing to facilitate community-engaged research. Outside work, I enjoy rock climbing, social dancing, and crochet.

2) Tell us about your dissertation: You crafted a project that included both a written treatise on accessibility and disability studies and a software application that you prototyped, tested, and refined over the course of a year or so. How did that mix come about?

In the first year of my PhD, I was looking for a Disability Studies community, because I didn't easily find that in my Computer Science and Communication Studies departments. I searched online for meetups and found the Disability Studies reading group at the University of Chicago. The group included academics across the humanities, social sciences, and STEM who were to engage with Disability Studies and discuss readings by Alison Kafer, Leah Lakshmi Piepzna-Samarasinha, Eli Clare, and more disability scholars, authors, and activists. It was great to be in a space where nearly everyone was coming from a place of critical disability politics and many of us identified as disabled.

One person in that group was a dyslexic anthropologist, Alex (pseudonym), who eventually became a research collaborator. In one of our first conversations, he told me spell checkers consistently do not work for him as a dyslexic social scientist. Alex was the one who initially put writing tool accessibility issues on my radar, pointing out many things that I took for granted about spell checkers, cognitive abilities, and language norms.

After wrapping up my first-year project, I looked for potential collaborators for my next research study. I went back to Alex and said "Hey, I've been interested in doing a project with dyslexic adults, spell checkers, and academic writing practices ever since you mentioned it in Disability Studies reading group. I think this is an important topic that computer scientists can learn from and do something about. I’d love for you to be involved." He has been supportive of this research arc since the get-go.

Sara: It’s so instructive, I think, for grad students to hear this—that ideas can come from lots of places, and that it’s worth it to be brave with a little seedling of an idea. So then how did it come together?

The first phase of my dissertation was interviewing dyslexic students across different fields about their experiences with academic writing practices and tools. The interviews also included a think-aloud activity where my informants walked me through their user experiences with spell checkers and other interfaces. This helped me get a more contextualized understanding of the accessibility issues before I started brainstorming and building anything.

I learned a ton about not only the technological shortcomings of existing tools, but also how dyslexic writers developed creative workarounds when the technology didn't work for them. For example, one issue participants mentioned was that spell checker interfaces can be illegible when they show a list of word suggestions that all look similar, like “referred, refeed, referend, revered.” Their workaround for these situations was to copy-paste options into search engines because the search engine results would show more context to differentiate different word options. Understanding both prongs—the accessibility design issues and writers’ workarounds in the meantime—generated insights on how the spell checker itself could be redesigned to be more accessible.

Sara: Yes—I love this. I think too many young technologists and designers might skip over the attentive, ethnographic-style interview that makes visible these deeply adaptive behaviors that are already present. It makes your design work so much more robust. And then you turned that research into a pragmatic technological tool, right? 

Yes! My interviewees’ adaptations informed what I did and didn’t build. I developed Jargon Manager, a software toolkit with a browser extension to opportunistically save terms while reading and a word processor extension that provides a custom autocomplete interface based on the previously saved content. The focus of the project wasn't to make a better algorithmic proofreader, but rather to increase the tailorability of text entry interfaces. There’s no such thing as a one-size-fits-all writing tool, so the data in each writer’s Jargon Manager is going to be unique. The technological contribution is the stronger integration of a writer’s local context and their tools, which goes beyond what’s currently available with the “Add to dictionary” button in spell checkers. 

The final part of my dissertation was deploying the Jargon Manager toolkit with Alex, the colleague who sparked my initial curiosity in this topic. We did a design-for-one assessment where Alex used the tool over seven months for writing his dissertation chapters. We had monthly check-ins to discuss: “How’s the writing process going? How is this toolkit helping? How is it not helping? How can we make it better?” 

One of the most surprising findings was how Alex reappropriated Jargon Manager for both multilingual and academic writing. His dissertation is based on fieldwork he's done overseas in Vietnam, meaning that he uses a unique combination of Vietnamese terms and anthropology jargon in his papers. He told me that spell checkers have not been helpful at all when he blends languages, so he appreciated the additional customization and flexibility of Jargon Manager for his academic writing. I loved seeing how Alex used the toolkit in unexpected ways.

Sara: So interesting. And then—tell us some of the challenges along the way. A PhD is a marathon, not a sprint, and there are always bumps in the road.

Absolutely. I’ll mention a bit about my positionality before I explain the challenges. This is a project that exists in the shared space between me and the dyslexic community. I am neurodivergent, but I'm not dyslexic. I cannot single-handedly create a tool that caters to dyslexic people.

I aim to be a responsible collaborator who amplifies what the dyslexic community wants and integrates that with prototyping. Also, I’m not trying to solve or cure dyslexia, because that's not actually what my dyslexic collaborators said they want. What they want is tools that are built with them in mind and for non-dyslexic people to be more understanding of neurodiversity and less ableist.

One challenge in my work is reconstructing researcher-participant relationships. Much of academia is historically set up for researchers to work with participants in anonymized, controlled lab settings. It often goes like this: participants are brought into the lab and answer a scripted list of questions, or researchers record them as they follow instructions, and then they are paid and leave. But that's not going to fly in a disability and design process. If I work on my own in a vacuum and don’t meet with the community until the later stages of data collection, then there's a very high potential for me to do harm inadvertently, despite my good intentions.

Instead, I meet members of the disabled community and do qualitative research before designing and building anything. I reach out to ask disabled communities to share their expertise and walk me through how they use technology, and then I put that in conversation with technology design and development. For example: Here are what the accessibility issues are that we found out by talking to dyslexic people, and also as a builder-engineer speaking: Here's what we could do to improve the technology. 

Another challenge is scoping meaningful outputs for both design collaborators and academics, as well as considering the sustainability of design interventions. This can be tricky because a norm in academia is for researchers to end their involvement with a community once the research is published or another milestone is achieved. However, these projects are not just about academic career advancement; there are social stakes.

If I engage in an iterative design process with marginalized communities and claim research contributions, I am also responsible for giving back to those communities.

How can I use my platform to both fulfill my goals as an academic and support the collaborators who were willing to share their time and perspectives to make this project possible? Co-authoring publications could be desirable for those who prefer recognition over being an anonymous participant, but if collaborators are not academics, they likely want different products and impacts. Throughout my projects, I check in with collaborators about what short-term and long-term outputs are meaningful to them, which may include a tech product, media, documentation, events, or some combination. 

Regarding sustainability: when researchers co-design and deploy interventions with a community, it’s important to consider the upkeep beyond the research study period. Academics eventually switch their focus to different research questions and projects, while collaborators may want to continue using the co-designed products. How does the community want to access the prototype for the foreseeable future, as well as modify it themselves when their circumstances change? If the intervention was a series of workshops or other events, does the community want to facilitate that multiple times beyond the initial research study timeframe? Again, I recommend checking in with collaborators and adapting plans to be on the same page about timeline, involvement, and long-term access to the designs as desired.

3) One of the things that’s so impressive to me about your work is your determination to a) dive deeply into disability studies, especially to understand the many pitfalls of technology when it’s created in a vacuum, and b) build from those very convictions. Thinking and making, together. I often speak about this as the wedding together of critique and repair—a set of concepts that are easy to affirm but hard to hold together in practice. How do you think about your stance between and among several fields of research?

I think a lot about what it means to build bridges between Computer Science and Disability Studies, because STEM historically does not center disability and access.

Painting with a broad brush here: a big part of Disability Studies is in valuing disabled perspectives and pointing out instances of ableism throughout our society, whether that's enacted through institutions, policies, healthcare, media, devices, and the built environment.

On the other hand, technology disciplines design and build software and hardware that often end up as black boxes. Despite designers’ and engineers’ good intentions, technologies perpetuate many of the systemic problems and cultural biases that Disability Studies critiques through humanist disciplines.

If you put these bodies of work and skillsets in conversation with each other, you realize there's a lot we could do together beyond what either discipline could alone. We need to be able to name and understand the problems that ableism causes, and we also need an iterative design process to build anti-ableist interventions that reduce harms in the status quo. That involves learning how to navigate current systems as well as dream up and design better ones in solidarity with marginalized communities.

I start my research process not from the literature or research gaps about disability and tech, but with relationship-building and learning about different communities’ lived experiences. When I connect with people and they find out I build software and hardware, it often segues into us chatting about technology and accessibility in everyday life. Ultimately we’re trying to get on the same page about how centering disabled people’s experiences reveals what could be designed better moving forward. We saw this in the dyslexic writers study where my informants brought up not only the issues with spell checkers, but also university accommodations and society’s expectations for perfect spelling and grammar. This is where critique and repair starts: “How do technologies and social norms perpetuate marginalization? How are disabled communities already developing their own adaptations and workarounds to the accessibility barriers? How might we reimagine technology and social norms to expand the possibilities for participation? What will we build and not build?" Over time we develop a canon of design commentary and possibilities that can evolve into a research proposal.

I love my interdisciplinary training. I end up doing a bit of everything, which includes diving into critical humanities and computing theories, going out into the field to get a contextualized understanding, coming back to the design studio for prototyping, and returning to the field to test how my prototype works in the world. I find a lot of satisfaction in following through with that iterative process, though I have to set up additional structures and support systems to make it work. But I wouldn’t choose to have it another way at this point in my career.

4) Tell us about selecting a graduate school program. How did you think about this while at Olin, and what advice would you offer to undergrads thinking about a grad school career in general?

At Olin, I was figuring things out in summer 2015, before senior fall. My initial plan was to apply to both grad school and industry and see what happens. It ended up being overwhelming trying to do both of those job application tracks at the same time. I started with academia because the deadlines are once a year in December and January. At that point I had done summer research at Olin, a software internship in industry, and an interdisciplinary fellowship program at the Open Style Lab. The constellation of three different work experiences helped me figure out what I wanted to do next.

Whether it be in academia or industry, I knew I wanted a role where I could work in a community-engaged design process focused on disabled communities and accessibility. I wanted to use the methodologies introduced in Olin’s human centered design courses, as well as build things and deploy them out into the world. That whole loop. I talked with different professors at Olin—including you—about what I could do. One suggestion I received was to spearhead my own research agenda where I could set up that loop as a PhD student. So that helped me orient myself with some criteria: looking for interdisciplinary programs, wanting to extend my training as both a social scientist and designer-engineer. It was also important to find an advisor, or multiple advisors, who could support both the computer science and qualitative research threads of my work. The context and type of technology I'd work on was relatively flexible.

To find potential advisors, I did a literature search for professors doing accessibility and human-computer interaction research. I read some of their research papers to know the methodologies and communities different faculty had worked with, which helped me get a sense of how likely they could support the projects and approaches I wanted to do.

After identifying a handful of potential advisors at different institutions, I reached out individually to ask whether they would be accepting new students and if they would be willing to chat to confirm shared research interests before I applied to PhD programs. I will also amplify the advice I was given, which is to make sure there are multiple faculty in the department you could collaborate with, because you will be required to have multiple faculty from your department on a dissertation committee. I eventually found my PhD advisor, Anne Marie Piper, at Northwestern – she is both a social scientist and designer who seeks out perspectives from Disability Studies. Northwestern is in the Chicagoland area, which has great public transit and many options for finding hobbies and community outside of work, which I also wanted.

Some advice that I could have benefited from and did not get at the time was about the reality that a PhD program is a huge investment. PhD students are paid a stipend, but it's different compared to the salary in an entry-level tech industry position. You owe it to yourself to do some low stakes experimentation and career exploration…and not just go into graduate school for the sake of more school or postponing going on the job market. It's really about figuring out the itch you're trying to scratch. Is that only possible in graduate school, and is it worth all the uprooting and precarity of being a student for 4-6 more years? It's going to be very different from undergrad because PhD students are often left to their own devices. You have to be very self directed. There aren’t many guardrails. Even if you think you are prepared for it and even if you have been very good in academic environments so far, the structures and incentives of a PhD program are different enough that's not always a good indicator.

Below is a non-exhaustive list of questions to ask yourself and low stakes experiments to start:

• How is a PhD going to be a stepping stone for what you want to do next? Being a graduate student is a transient position. What career roles are you considering? Do the programs you’re considering have alumni who advance into roles that you’re excited about?
• What research activities are you most excited about? Gather some information about the research practices in your subdiscipline of interest. Are researchers working in the lab or out in the field? Do the research papers predominantly feature quantitative, qualitative, and/or mixed methods? Are there already researchers exploring interdisciplinary topics and approaches? Maybe you know this from undergraduate research experiences, otherwise you can look at the methodologies and theories mentioned in publications. For example, I knew I wanted to do both qualitative research and design projects with disabled communities. The goal is to be explicit about what would make research projects fulfilling to you and find a graduate program that supports most, if not all, of your criteria. 

• Is your interest in an academic sub-area enough to sustain you through the marathon of reading, writing, proposing a study, navigating bureaucracy, collecting data, analyzing that data, and presenting it? Building on the previous question, you might not get to do the parts you’re most excited about very often. Search online to find a syllabus for a graduate seminar in the subdiscipline that you're curious about. Try some of the readings and assignments to see if you are keen to read in the genre, engage in academic work practices, and join the dialogue as a researcher.

• Are you excited about teaching and STEM education? A PhD program doesn't set you up to get training in inclusive pedagogy unless you go out of your way for it. If you want to scratch that itch, try volunteering as an instructor or mentor at the local library or school district. Depending on the field, you could also experiment with being an adjunct instructor for a university and teaching a class in your specialty.
• Do you want to do community engagement? Academia does not incentivize everyone to do community engagement by default. If you want to work with your local community on a design project, that's something you could pursue in different forms of volunteering and outreach. You don't have to be a PhD student to get started with that.

5) Tell us about your formative experiences as a young person, including at Olin, that set up the interests that you have today. It’s a myth that the march of a life is a steady, even progression, but how does hindsight help you make sense of your story so far?

Thanks so much for this question! Thinking back on my Olin and Northwestern timeline: the highlights are summer research, internships, and classes when I dove into assistive/adaptive technology design and qualitative methods.

In the summer after my sophomore year, I worked on Olin summer research with Paul Ruvolo. It was the beginnings of his lab's project about accessibility and blind grocery shoppers. A few other Oliners and I took the research idea and ran with it. We spent half our time meeting members of the visually impaired community in the Greater Boston Area, learning first-hand about how blind people use technologies in everyday life. The other half of our time was spent getting our feet wet with computer vision libraries in Python and learning how to read Computer Science research papers. The part I worked on eventually became a poster I presented at the Grace Hopper Celebration.

In retrospect, that summer was an initial exposure to research workflows I still practice now: going into the field to meet disabled communities, academic reading, using design techniques from Collaborative Design, and prototyping various software and hardware ideas. 

The following summer, I did a front end development and graphics internship at Onshape. That was a glimpse of what it might be like to be a programmer in industry. I enjoyed coding a bunch and the coworkers there were great. But that helped me realize I wanted to do more than just code in my job description. I also wanted to engage more with the design and narratives of the technology. 

Additionally, that summer I was an engineering fellow at the Open Style Lab, which I found about from the Adaptive/Assistive listserv at Olin. Open Style Lab puts together teams of an engineer, a fashion designer, an occupational therapist, and a member of a local disabled community. Everyone works together on an adaptive fashion project that's situated in that disabled collaborator's context. This was a great foundation and formative experience—like, "Wow, okay, what does it mean to have a truly interdisciplinary collaboration where we're all coming from different backgrounds and speaking different languages? How do we meet each other halfway to think together about assistive/adaptive design?”

My Open Style Lab teammates and I had more commonalities than I expected given the different disciplinary and professional labels. Olin exposes students to many prototyping and fabrication techniques. Occupational therapists teach and create adaptations based on their clients’ context. Fashion designers are skilled prototypers of garments. We have this shared building and creative practice. 

It was also great to have that project be a design journey in which our disabled client was there from the very beginning. We engaged in an iterative process where the fashion designer, occupational therapist, and I would do a bunch of brainstorming, prototyping adaptive apparel, and getting feedback from our client over ten weeks. This experience was what sold me on a longitudinal design-for-one, which I eventually did a variation of for my dissertation. 

The last experience I'll mention from undergrad was when a couple other Oliners and I took an independent study with Mel Chua, who's a ‘07 Olin alum. She was doing a postdoc at Olin at the time, and we did an independent study called Qualitative Methods in Practice. That was a deeper dive into interviews and observations beyond the primer from Collaborative Design – ultimately getting more hands-on experience with what it means to be an interpretive social scientist and that really made me think: "Oh, I want to do more of this in grad school."