The UI designer must be able to anticipate cognitive difficulties of users in the UI design process. However, the designer is likely to make erroneous judgments in the context of increasing functionality. Furthermore, time constraints in the development process exacerbate the design problem. There are various techniques to support the UI designer in the design process, including abstract design principles, specific design guidelines, design cases, design inspections, and design metrics. Metrics can summarize the status of a UI design solution more objectively and more accurately than human designers. This paper aims to develop quantitative metrics based on a unified framework for interaction design, which decomposes UI design problem into the four components: information architecture, task procedure, system dynamics, and physical interface. Three metrics were proposed to assist designer"s decision-making, including incongruity, complexity, and inefficiency. A case study shows that the proposed metrics can support the designer"s decision making in an efficient manner.