Human agency is our capacity to make choices and exercise control. My research focuses on how software systems can both inhibit and support the agency of users. My projects involve studying and building technological interventions that enable users to become more agentic actors. These agentic actions and capabilities are always situated within the personal, professional, or community context that we are engaged in. I believe that by better understanding how our interfaces impact each individual stakeholders’ agency, we can create systems that will help their users to have more meaningful and productive interactions and relationships. Through exercising agency within these systems, other stakeholders are more able to see their value and contributions. A few example projects that I have worked or am working on are below.
Exploring Agency Over One’s Attention in Email
In this project I was interested in how users were Getting Lost in Email. I outlined how email can draw its users down a path of interaction, almost against their will. In my work, I used a novel methodology to study email usage and made several contributions to the field. For example, I was the first to apply Conversation Analysis to email interactions. Specifically, I outlined how the lack of a clear summons mechanism for conversational emails was problematic. Ultimately I found that my participants were Getting Lost in Email in 23% of their sessions.
Supporting Crowd Workers’ Agency in the Labor Market
While a Research Scientist at Xerox Research Centre Europe, investigated the practices of workers on the Amazon Mechanical Turk (AMT) crowdsourcing platform. In this work I described and contrasted the experience of Being a Turker in the United States and India, and outlined the invisibility of their work products and practices to the various stakeholders. Particularly, I refuted the notion that workers on AMT were working for fun, and outlined how AMT served as a safety net. Based on these findings I constructed the TurkBench tool, that aimed to increase the agency of workers by re-rendering the presentation of information and imbalance of functionality on the platform.
Supporting Community Co-production
I am working to better support agentic actions for members of small, local communities. As part of this work I have built WithShare, a mobile application that embodies the concept of co-production. In this case, the concept of co-production helps to desconstruct the larger goal of stronger, more connected communities into more concrete component activities. Within these activities, all stakeholders are recognized as vital to jointly enacting the activity. This approach helps to make an amorphous, intractable goal like a more connected community into a concrete, tractable activity between individuals. Along these lines, WithShare encourages community members to enact activities with other community members that they may not know. These interactions are the components for building a stronger, more connected community.