Meta-design extends the boundaries of system design from the creation of one original system to an ongoing co-design process. This paper describes the beginnings of two experiments exploring the effect of participation and meta-design guidelines on idea generation, to illustrate how meta-design concepts could be realized in design practices.

We describe Herbal, a high-level behavior representation language that compiles into Soar, ACT-R, and Jess. We present a large model (541 rules, one of the largest) we built one afternoon. Its fit to learning data suggests that Herbal helps create models 10x faster than working directly in a cognitive architecture.

We present YouTube as a case study that exemplifies the opportunities and challenges to ethnographic HCI research in large-scale online social spaces. We find that the constant change of users, content and tools that typifies large scale online spaces calls for renewed thinking of the ways in which HCI-related ethnography is conducted in them.

Virtual world environments afford new possibilities for design in the field of software engineering and in HCI in general. Virtual worlds provide specific advantages, such as interactive awareness, and allow us to explore and advance design theories like meta-design.

There is a need in HCI to study how issues of trust and privacy can and do affect the ad hoc negotiation of security rules and how they are managed by humans in actual practice. In this paper we present a field study of security and privacy through the use of interviews and observations that examine the physical and electronic security practices of childcares and medical offices. We show that the issues of human-mediated access management, information duplications, and the creation of a community of trust all affect aspects of the human-side of security.

There are impactful programs, such as handwriting recognition, that cannot be written by explicitly specifying behavior. Statistical machine learning algorithms are effective tools for solving these problems but are difficult for developers to use. We discuss difficulties faced by developers using machine learning and future work aimed at supporting developers.

The rise of massively distributed systems for human knowledge and cognition has led to an emerging form of distributed human computation. In this paper we describe our current work exploring the capabilities of distributed, ad-hoc human computation with a simple, classical computational problem: playing a game of chess.

Current social network visualization tools are tool-rich but process-poor: they simply offer investigators a collection of analytical tools and visual outputs, but ignore the sensemaking process in social network analysis. This study proposes a sensemaking framework of social network analysis supported with visualization tools. With this framework, we implemented a prototype system, VisNetSense, and conducted preliminary user evaluation.

This paper describes our efforts at designing and evaluating a shared research experience for undergraduate students centered on a common book through an approach and analysis centered on the use of a common book. We present our initial efforts at moving from a co-located to a distributed experience through the use of Facebook, Google Groups, and other technologies.

As touch screens become increasingly pervasive, most touch screens remain partially or completely inaccessible to people with visual impairments. This paper describes our efforts to improve current accessible touch screen interfaces through participatory design and introduces a technique to make any touch screen accessible using computer vision-based gesture tracking.

We examine two instances of popular social games on Facebook, to understand the social interactions and influence dynamics in the invitation/adoption networks. We investigate what factors makes more influential structures, how influentials are distributed, and how the games copy social networks from Facebook and exert network externality over time.

We investigate how mental models of a photo sharing website are shaped by observing behavior of existing users. Despite maintaining uniform site content and design, mental models are significantly influenced by social cues embedded in existing content highlighting behavior, manifesting different behavioral explanations, audience perceptions, and predictions of unseen features.

Prototyping is central to learning in design. How does creating and receiving feedback on multiple prototypes in parallel ' as opposed to serially ' affect learning, self-efficacy, and design results? A design experiment found that a parallel approach leads to more divergent ideation, more explicit comparison, less investment in a single concept, and better overall design performance.

This research uses patients' psychosocial information as a lens to examine medical practice pertaining to information creation, sharing, and documentation. It employs ethnographic observations and in-depth analysis of medical records. It reveals clinicians' complicated views about psychosocial information and argues for considering medical records as both representations of medical work and of patient.

Online payment services have allowed us to leverage financial incentives in many virtual interpersonal interactions, but are there drawbacks? In this work, I found that the use of financial incentives can undermine social relationships by reducing the social interactions and changing the expected norms of interaction from communal to exchange-based.

Applying educational games to the training of social skills by simulating human behavior with embodied conversational agents is an emerging topic within HCI. We show, through the use of a culturally-grounded virtual environment, that participants who approach the simulation as a social interaction improve their cross-cultural negotiation skills through gameplay.

As use contexts become increasingly complex, designing for usability may no longer be sufficient. While designing useful products has always been at the core of HCI, there is little understanding about what the term 'usefulness' means. This research addresses this issue by exploring the relationship between usefulness, context, and usability.

Humans and machines have competing strengths for tasks such as natural language processing and image understanding. Whereas humans do these things naturally and with high accuracy, machines offer greater speed and flexibility. CrowdFlow is our emerging human-machine computing model for blending the two for tighter control over the speed-cost-quality tradeoff.

Social software can help connect patients with their friends and family, with other patients, and with caregivers, giving them access to information and social influence. Through the development of two applications, we are identifying opportunities and challenges for building applications as part of existing social network sites that can leverage a patient's existing relationships.

Ever wonder why the internet is slowing down in your home? Or who's hogging the bandwidth? In this boaster paper, we describe our ongoing field trial of Kermit, a visual home network probe that enables households to answer these questions.

This paper argues that design rationale has become irrelevant by effectively ignoring the practical contexts for which it was originally developed. We propose the adoption of a phronetic research agenda that supports the study of human value rationality in design. Through a Participatory Design of Design and similar efforts, we can make design rationale matter again.

This extended abstract describes initial work that we are conducting regarding the adoption and appropriation of a SNS deployed in a global enterprise across different geographic regions and organizational function. We use this analysis as a first step in identifying specific local and regional practices in social software to help designers and researchers create systems that mitigate negative impression formation.