Externalizing Thought: Improving the design of software through diagrams

专家的设计比新手的好，但原因尚不清楚。在这里，我们提供了理解软件设计专业知识及其发展的新方法。这些方法涉及分析图表来揭示设计师的认知结构，已经在其他依赖草图和可视化来实现和微调设计的学科中开发出来。它们还没有被应用到软件密集型系统的设计中。该项目的一个独特而有价值的方面是，它嵌入在设计课程中，这是一个学习设计的机会，同时也提高了设计教学。该项目还将研究在该领域实践的专家设计师。我们的努力集中在以下广泛的研究问题上：专家级软件设计师与新手的区别是什么？专家和新手如何表示问题和解决方案？哪种教学技巧在提高新手技能方面最成功？最重要的是，我们如何提高设计能力？糟糕的设计以及由此产生的糟糕的软件阻碍了许多政府和企业机构的协调。设计师教育的任何改善都可能通过他们在实践中的努力和在职业生涯中指导其他设计师来放大。具体的影响应包括改善软件密集型系统设计教育，从而提高软件和社会系统的效率，由本和其他设计方案的毕业生创建。因此，可能的影响包括更好的国土安全，更好的经济增长，更好的社会服务，以及改进系统设计的其他影响。因此，在这一领域的见解可以帮助我们改善广泛的人类活动。

Designs of experts are better than those of novices, but it is not clear why. Here, we provide new methods for understanding software design expertise and how it develops. The methods, which involve analyzing diagrams to reveal the cognitive structures of designers, have been developed in other disciplines that rely on sketches and visualizations to realize and fine-tune design. They have not yet been applied to the design of software-intensive systems. A unique and valuable aspect of the project is that it is embedded in a course in design, an opportunity to study design as it happens and to improve design instruction at the same time. The project will also study expert designers practicing in the field. Our efforts focus on the following broad research questions: What differentiates expert software designers from novices? How do experts and novices represent problems and solutions? Which teaching techniques are most successful in increasing the skills of novices? And, most generally, how can we improve our capacity to design? Bad design, and the resulting bad software, impedes coordination in many government and corporate institutions. Any improvement in the education of designers may be amplified by their efforts in practice and in mentoring other designers over the length of their careers. Specific impacts should include improving software-intensive systems design education, consequently improving the efficacy of software and social systems created by graduates of this and other design programs. For this reason, possible impacts include better homeland security, better economic growth, better social services, and other effects of improved systems design. Thus, insights in this domain may help us improve a broad range of human activities.