Visual Programming Labs for Teaching Computer Science Concepts to Undergraduates

ACM计算机科学核心任务组和NSF本科教育学科研讨会都建议为学习计算机科学的本科生安排结构化的实验课程。根据我们的经验，当学生使用传统的编程语言时，他们在实验室课程中几乎没有进步。因此，我们正在为一系列“可视化实验室”创建实验软件和手册，在这些实验室中，学生使用可视化编程环境而不是传统的编程语言。他们创建和测试模型（例如CPU、机器人的传感器配置、有限状态机、数据库查询或遗传算法），而不是编写代码。每组实验都是基于Task Force课程中的一个知识领域。我们之前已经为计算机体系结构、编程语言和机器人技术创建了实验室。我们现在正在为数据库和信息检索以及另外两个知识领域（暂定为人工智能和软件方法论和工程）创建实验室。每一套实验室都是通过以下方式创建的：定义教学目标；设计用户交互；制定软件的规格和设计；使用C语言和EYES可视化编程工具包实现软件；编写和桌面出版手册；对软件和手册进行用户测试；并根据用户测试结果对软件和手册进行修改。然后在课堂上使用软件和手册。我们之前的研究表明，与没有实验室相比，视觉实验室能提高学生的表现。我们目前正在进行正式的研究，将涵盖相同概念的可视化实验室与传统编程实验室进行比较。我们使用的实验设计是与该大学的人体实验委员会共同开发的。这些研究将揭示可视化实验室是否是本科计算机科学实验课程的有用辅助，并将产生关于可视化编程对教学有用性的信息。研究结果将发表在期刊和会议论文上。实验室软件和手册可分发给其他感兴趣的学院和大学。

9345708 Williams Both the ACM Task Force on the Core of Computer Science and the NSF Disciplinary Workshops on Undergraduate Education recommend scheduled, structured laboratory sessions for undergraduates studying computer science. In our experience, students can make little progress during a lab session when they are working in a traditional programming language. Therefore, we are creating lab software and manuals for a series of "visual labs" in which students use a visual programming environment instead of a traditional programming language. They create and test models (of a CPU, a robot's sensor configuration, a finite state machine , a database query, or a genetic algorithm, for example), rather that write code. Each set of labs is based on one knowledge area from the Task Force curriculum. We have previously created labs for computer architecture, programming languages, and robotics. We are now creating labs for database and information retrieval and for two other knowledge areas (tentatively artificial intelligence and software methodology and engineering). Each set of labs is created by: defining pedagogical goals; designing user interaction; developing specifications and designs for the software; implementing the software in the C programming language and in the EYES visual programming toolkit; writing and desktop publishing a manual; conducting user tests of the software and manual; and revising the software and manual based on the results of the user tests. The software and manuals are then used in class. Our previous studies revealed that visual labs improve student performance, compared to no lab at all. We are currently conducting formal studies to compare visual labs to traditional programming labs that cover the same concepts. We are using an experimental design developed in conjunction with the university's human subjects committee. These studies will reveal whether visual labs are a useful adjunct to the undergraduate computer scie nce laboratory curriculum and will yield information about the usefulness of visual programming for instruction. Results will be published in journal and conference papers. The lab software and manuals are available for distribution to other interested colleges and universities.