Increasingly, designers seek feedback on their designs from crowd platforms such as social networks, Web forums, and paid task markets which demand different amounts of social capital, financial resources, and time. Yet it is unknown how the choice of crowd platform affects feedback generation. We conducted an online study where designers created initial designs and revised the designs based on crowd feedback. We measured the quantity, quality, and content of the feedback received at two iterations and from crowds driven by social status, enjoyment, and financial gain. Our results show, for example, that task markets yield more suggestions, online forums provide more process feedback, and social networks give the most suggestions without payment. We contribute an emergent framework for crowd feedback selection, opportunities for enhancing feedback services, and an experimental platform that researchers can adapt to reduce the burden of conducting online studies of design feedback.
The number of hackathon events worldwide has nearly quadrupled in the last five years. Despite exponential growth across diverse industries and increasing interest across academic disciplines, our integrated understanding of the phenomena of hackathons is limited. We conduct the first multidisciplinary literature review of publications from 1999 to 2022 to understand the conceptualization of the phenomena over time. We find that hackathon research can be categorized into 4 core areas (purpose, format, processes, and outcomes). Research was first driven by a purpose (innovation, learning, and collaboration), followed by an examination of how formats adjust to purpose to influence what happens (processes) and what is produced (outcomes), and critical reviews of the hackathon phenomena. We contribute a unifying framework with these four core areas to inform future directions of hackathon research and practice, as well as a discussion of the need for longitudinal and multidisciplinary research of hackathons.
14 April 2023 in Proceedings of the ACM on Human-Computer Interaction
One-to-many coaching is a common, yet difficult, coaching technique used in environments with many novices learning to solve ill-defined problems. Intelligent systems might be designed to support 1-to-many coaching but designing such systems requires a 1-to-many coaching model that details novices' challenges, coaches' strategies, and coaches' goals. To build such a model, we conducted interaction analysis on 24 1-to-many coaching sessions with novices developing new products in a university incubator and conducted retrospective analyses with 3 coaches and 30 novices. We contribute a model that demonstrates that coaches in a 1-to-many setting not only need to help novices develop metacognitive skills (just as in 1-to-1 coaching), but also need to utilize the presence and expertise of a group of novices to learn from each other, to mitigate their fear of failures, and provide them accountability. Our model informs design implications for future intelligent coaching systems to (1) assist coaches in monitoring and comparing many novices' progress, learning, and expertise; (2) provide novices with checklists, templates, and scaffolds to help them self-evaluate, seek-help, and summarize learning; (3) showcase failures and growth; and (4) publicize planning and progress to provide accountability.
16 October 2025 in Proceedings of the ACM on Human-Computer Interaction
Entrepreneurship requires navigating open-ended, ill-defined problems: identifying risks, challenging assumptions, and making strategic decisions under deep uncertainty. Novice founders often struggle with these metacognitive demands, while mentors face limited time and visibility to provide tailored support. We present a human-AI coaching system that combines a domain-specific cognitive model of entrepreneurial risk with a large language model (LLM) to proactively scaffold both novice and mentor thinking. The system proactively poses diagnostic questions that challenge novices' thinking and helps both novices and mentors plan for more focused and emotionally attuned meetings. Critically, mentors can inspect and modify the underlying cognitive model, shaping the logic of the system to reflect their evolving needs. Through an exploratory field deployment, we found that using the system supported novice metacognition, reduced mentors' cognitive load, and improved meeting depth, intentionality, and focus--while also surfaced key tensions around trust, misdiagnosis, and expectations of AI. We contribute design principles for proactive AI systems that scaffold metacognition and human-human collaboration in complex, ill-defined domains, offering implications for similar domains like healthcare, education, and knowledge work.