Supporting Collaborative Design through Risk Analysis: Benefits of Calculated Risk in the Design of Interactive Systems

As software systems continue to grow, and as project teams become larger and more distributed, support for project management in collaborative environments is critical. Management tasks include maintaining team coordination, monitoring progress, and, of particular interest for this work, managing risk tasks often add significant overhead to a project. To reduce overhead, management tasks must be integrated, whenever possible, directly into the software design and development process. Additionally, to prevent common problems from reoccurring in different projects, developers must focus on reusing the knowledge gained and the lessons learned through previous projects to guide future endeavors.

The overall goals of the work contained within this thesis are to define reusable, project-related knowledge as project risks and to utilize that knowledge in the development of a risk-driven management model to be integrated within a human-computer interaction (HCI) design process. Existing risk management techniques typically involve process-related knowledge, such as project planning and client involvement. However, HCI as a discipline is more concerned with product-related design knowledge. Claims structure product-related knowledge for reuse by explicitly stating the positive and negative tradeoffs of incorporating a particular feature in the design of a system. By managing these negative tradeoffs as design risks, HCI designers can identify and focus on the most critical design issues throughout the course of a project. This systematic approach to solving design issues helps to ensure that designers make informed design decisions rather than following an ad hoc design process.

Building upon existing risk management techniques from other domains, this thesis delivers a risk-driven, claims-based management model for HCI design. In doing so, this work transfers techniques traditionally used in managing process-related knowledge into a new domain for use in managing product-related design risks. The need for risk management in software design is argued through a review of existing collaborative tools, resulting in a series of guidelines for providing project management support. An initial risk model is then presented, along with the results of a user evaluation conducted to determine not only the accuracy of risk prioritization, but also the overall benefit of applying risk management within the context of HCI design. Following a discussion of these results, several directions for future work are mentioned both to further the quest for a true design science and to improve the standards by which software projects are managed.