Overview of Current Research

Data use in Advocacy
Nonprofit advocacy organizations have some of the most powerful voices in contemporary politics. Nonprofits are also responsible for informing governmental decision making around allocating resources to address needs in our society. Our research combines qualitative studies of the various genres of data work in advocacy organizations, finding powerful exemplars of data feminist principles in action, with content analyses of the rhetorical data work of political action committees on social media.

• Data Rhetoric in Electoral Communications [CSCW'23]
• Considerations toward Enacting Equitable Data Work [CSCW'23]

Interorganizational Data Infrastructures for Immigration
Collaborative data work among the interagency network of organizations supporting migrants, refugees, and asylum seekers is notoriously difficult as government institutions, nonprofit organizations, and private entities try to balance refugee-client privacy with the flexible access needed to coordinate across a dynamic network of organizations. Numerous government agencies, nonprofit shelters, legal service providers, and transportation providers each have unique data infrastructures, collecting different kinds of data about asylum seekers and refugees. At the same time, the southern U.S. border is a focal point for (inter)national politics and immigration policy. Data collected about asylum seekers influence organizational work, local and national policy, as well as fuels many media narratives and news coverage. Since August 2022, I have collected data via 300+ hours of participant observation and semi-structured interviews with many organizational stakeholders serving migrant communities.

Toward More Ethical use of AI in Immigration at the Southern Border
As asylum seekers from different countries of origin enter the U.S. via the southern border, they must transition to using new information systems. Despite the many barriers (e.g., linguistic, socioeconomic, legal, etc.), not only must asylum seekers use these systems, but their system use can have serious implications for their status in the U.S. (e.g. leading to consequences like detention and deportation). Many of the technologies deployed by both government institutions and private contractors rely on AI-based technologies such as NLP analysis and deep learning for image classification. AI-driven solutions are prone to societal biases, especially when they are directed toward individuals coming from countries with relatively little data work. In such cases, AI systems are forced to rely on irrelevant proxies that can produce harmful recommendations for marginalized communities. This line of research seeks to understand (a) asylum seekers’ folk theories about the technologies deployed around them and systems that evaluate their data and information, (b) how folk theories cluster and (c) in/consistencies in the experiences of different groups of asylum seekers.