You can get more than you make

Clearly, exploring multiple alternatives can lead to better designs. Despite this wellknown fact, current computer-aided design systems provide only the most rudimentary tools for generating, storing, and visualizing alternatives. Hysterical space is a novel approach to discover alternatives in the solution space by using the interaction history with a parametric model. Implicit in any parametric model are the states a designer might have reached by combining variable settings in new ways. Such a model exhibits hysteresis, that is, path dependence — thus the name hysterical space. The thesis uses a simple definition of hysterical space as the Cartesian product of variable settings. It provides orderings of the space that yield feasible interactive search strategies. In turn, these orderings suggest interaction designs, which we report as working prototypes. Limited user evaluation supports a claim that hysterical space may be a useful approach to design space exploration.