Sears A L
Computer Science Department, University of Maryland, College Park 20742.
Int J Biomed Comput. 1994 Jan;34(1-4):149-57. doi: 10.1016/0020-7101(94)90018-3.
Automated metrics will allow user interfaces to be partially evaluated early in the development process. They allow potential problems to be identified and corrected prior to user testing, saving both time and money. As a result, improved interfaces can be developed without additional costs. Both task-independent and task-sensitive automated metrics appear promising. Task-independent metrics should prove most useful for predicting user preferences and somewhat useful for predicting user performance. Task-sensitive metrics should prove useful for predicting both user preferences and performance. Several metrics are introduced and research directions are discussed.
自动化指标将允许在开发过程的早期对用户界面进行部分评估。它们能在用户测试之前识别并纠正潜在问题，从而节省时间和金钱。因此，可以在不增加成本的情况下开发出改进的界面。与任务无关和与任务相关的自动化指标似乎都很有前景。与任务无关的指标在预测用户偏好方面应该最有用，在预测用户性能方面也有些用处。与任务相关的指标在预测用户偏好和性能方面都应该很有用。本文介绍了几个指标并讨论了研究方向。
