Research

My work has largely focused on measuring the extent to which sociotechnical systems are helping users to achieve their goals. The answer is never just technical — it requires deep understanding of users and their goals, and then operationalizing these into concrete metrics and statistical models. These themes trace that question across different domains, methods, and stages of my career.

Measurement as Design

Sociotechnical systems routinely optimize for metrics that are easy to measure but miss what actually matters. My work addresses this at two levels: ensuring we're measuring the right constructs — belonging rather than engagement, interaction quality rather than volume — and then operationalizing these constructs into trustworthy models and metrics that can be used at scale across organizations.

Online Communities

How do we know if an online community is effectively achieving its goals? My work examines how communities form, the motivations which drive their formation, how relationships develop within them, and a broader sense of community attachment evolves over time.

Moderation & Governance

Healthy communities require governance; in most community platforms these are led by volunteer moderators with varied backgrounds, skills, and goals. My research finds that effective moderation is less like policing and more like professional practice. Understanding moderators at a deeper level allows us to build more effective tools for them, translating into healthier, more effective communities.

Social Information Sharing

How do social signals shape the way that people find and make sense of information? My research has explored information overload in social streams, how social contexts shape search strategies, and how we can design tools to help individuals manage the flood of socially-mediated information.

Visualization & Representation

Analyzing complex social systems requires new methods for navigating large-scale network, text, and other kinds of data. Finding different ways to help analysts represent and interact with data can provide novel insights into the study of these systems.