Participatory Action Research

An AI for Good campaign I initiated in 2022 drew my attention to how emerging technologies shape disability rights. After finishing the contest and preparing to incubate related innovative ideas, I noticed that there are many cases of accessible technology in China, but do these innovations truly address the real problems? While researching existing technology solutions for the visually impaired, I came across several puzzling inventions, such as "shoes that do not distinguish between left and right", or "a seated treadmill". The imagination behind these innovations seems to view people with disabilities as abnormal, emphasizing their disability and suggesting a need for special care from the technology. How do these technological trends reflect societal views on visual impairment and other disability groups? How do they demonstrate disability-inclusive development within Chinese society? Is technology a promise or just a pretext?

These questions later grew into a series of interconnected studies on assistive technology patents in China. In collaboration with scholars from academic and disability communities, we analyzes more than 200,000 patent records from 1994 to 2023, examining how disability-related technologies are categorized, what kinds of needs and bodies they imagine, and which institutions drive their development.

Current outputs include two conference presentations at the 2024 International Communication Association (ICA) Conference and the 2025 Annual Meeting of the Society for Social Studies of Science (4S), both now being developed into journal articles, as well as one manuscript currently under review.

We have long been teaching children about digital safety, but who gets to define which risks are worth talking about? Although children's digital safety has become an important issue, knowledge about how rural children experience digital life under conditions of limited digital resources remains fragmented. At the same time, many rural areas in China are ethnic minority communities, where local gender norms tend to be relatively conservative. In such contexts, harms related to gender-based online violence may exist yet are often difficult to name directly.

Against this backdrop, I worked with Yuanzhong Family & Community Development Service Center, we take digital child safety as a broader framing and entry point. It draws on a three month digital safety curriculum and a three-day game-jam winter workshop conducted in a village primary school in Gansu Province, China, to examine how children, through discussion, co-creation, and prototyping, can participate in redefining which digital safety risk scenarios are worth narrating. These insights, together with broader cases of online gender-based violence, were translated into a gamified classroom toolkit for digital safety education. The toolkit is expected to be used in mandatory reporting trainings on child protection across different regions in mainland China.

I became involved in Open Source for Good after seeing how difficult it is for Tech for Development projects to survive beyond a single hackathon, team, or organization. Open source offered a possibility to document nonprofit needs, share technical lessons, and let ideas, code, and contributors continue as a relay rather than a one-off intervention.

I am honoured to facilitate OpenGood in KAIYUANSHE, a group originally formed in 2020 with the Wuhan2020 project in response to the COVID-19 pandemic. Since then, OpenGood has evolved into an ongoing effort to connect open source with social impact. In 2024, we officially launched an open call for OpenGood projects, encouraging open source initiatives in China that contribute to the Sustainable Development Goals. We also organize annual forums and continuously explore ways to support these projects and communities.

I was inspired by the concept of design justice, especially its emphasis on centering the communities most affected by design decisions. Although this experiment took place within a corporate setting, with clear limitations, it also had access to important resources: technical talent, budget, product infrastructure, and organizational attention. Under this scenario, how could marginalized communities leverage these resources? And how could a corporate initiative move closer to community-centered innovation?

To experiment with these question, the contest redesigned the position of visually impaired users by inviting them as experts, judges, and mentors. Together with community partners, we co-created a visual impairment needs ecology map and a set of inclusive design toolkits for engineering teams. Instead of asking teams to build quick demos in a traditional 24 or 48 hour hackathon, we extended the timeline to over six months, giving engineers more time to learn from community lived experience and adapt their accessibility solutions.

The contest attracted over 10,000 registrations and generated 64 internal proposals. But the more meaningful outcomes were not quantifiable or immediate. Three years later, one demo from the contest continued through open collaboration with external NGO partners and developed into a publicly testable app. Several engineers who participated in the contest also stayed connected with accessibility work after leaving Big Tech, with some even moving into disability communities and continuing to work on accessibility innovation.

Due to the lockdown, I was stuck in Chengdu when COVID-19 began. At that time, I was looking for a way to distract myself from anxiety. When I noticed that my undergraduate professor was leading an emergency research group focused on the analysis and prevention of COVID-19, I approached him and volunteered to join.

The early work focused on pandemic data analysis, including estimating the basic reproduction number, conducting contact diary surveys, and tracing transmission patterns. At that time, I was concurrently working on some gender issues. After establishing an overall background in public health prevention, I started to notice some missing perspectives in the research. Early pandemic reports suggested that men were more susceptible to COVID-19 based on confirmed case ratios. Yet community transmission data showed a different pattern: in family-clustered, confirmed female cases became more visible. Under lockdown conditions, where most community transmission occurred within households, this raised a gendered question: did women's care role increase their exposure risk? Starting from this contradiction, I designed a study to test the hypothesis and argue for gender perspectives in pandemic prevention policy.

As the pandemic gradually stabilized after 2022, I began to reflect on the catalytic effect COVID-19 had on accessibility: as lockdowns pushed public services and everyday life online, the exclusion of elder and disabled within digital systems became much harder to ignore. This pressure also contributed to stronger accessibility and age-friendly design requirements for major apps in China, pushing tech companies to respond. Yet this raised a further question: if accessibility advances mainly through crisis pressure and external regulation, how can it remain sustainable once that pressure fades? This reflection has become the basis of an ongoing book chapter on COVID-19, disability, and accessibility.

This line of work was one of my earliest research starting points. Coming from a computer science background, I first entered gender studies through computational social science: using quantitative methods to trace how gender inequality appears in labor markets and care roles. At the time, data felt like a way to make inequality more concrete - to show patterns that might otherwise remain anecdotal or invisible.

But working with gender-related data also made me aware of a more basic challenge: the data needed to ask these questions is often fragmented, missing, or not collected in ways that capture intersecting forms of inequality. I began organizing gender datasets and thinking about data not only as evidence, but as the condition that determines what can be seen in the first place.

This became my entry point into marginalization as an analytic lens: a way to understand how gender, disability, and other forms of exclusion are produced through labor, data, technology, and policy systems.

SeeForShe was the starting point for much of the work I have done since. It began with a feeling that something was wrong. A conversation with one of my undergraduate professors stayed with me: if you feel that something is wrong, how might you think beyond your own experience and imagine what could be changed at a community level?

That question shaped the early direction of the project. I wanted to create ways for more people to name, and discuss this discomfort together. This was also a moment when #METOO was making sexual harassment visible in public conversations around the world. Yet in many everyday spaces, there still remained an elephant in the room.

Games became one way to open that conversation. They made the topic easier to approach: interesting enough for people who thought it had nothing to do with them, and safe enough for others to begin speaking from experience.

As the inital gamification project grew, I also began to wonder: how couit could contribute to more systemic change? This question pushed the project forward to inclue the development of an open-source anti-sexual-harassment toolkit for workplaces. It also became one of the reasons for my later shift into corporate social responsibility, where I continued to iterate on tech for development and system-level approaches to social change.