CRII: CHS: Overcoming Novice Programmers' Misconceptions of Program Behavior

本研究的目的是调查如何编程工具可以增强，以帮助新手程序员了解他们的程序行为。近年来，对学习计算机编程感兴趣的人数急剧增加。然而，新手程序员面临的一个特别具有挑战性的障碍是他们经常对代码的行为方式抱有误解。新手有这个问题的一个关键原因是他们通常对程序正确性的定义与专业程序员不同。虽然当前的工具在修复语法错误方面提供了有用的反馈，但新手程序员通常在语义错误方面有相当大的困难，这些语义错误发生在程序编译时，但行为与程序员预期的不同。本研究将探讨工具设计，识别新手程序员的误解，纠正误解，并防止未来的误解。采取以用户为中心的设计方法，研究人员将分享一系列比较用户研究和现场部署的结果，以衡量工具设计的有效性。这些工具的可用性可以降低新手程序员面临的障碍，并提高学习如何编程的可访问性。特别是，该项目将开发新颖的工具设计，以实现三个相互支持的目标：（1）通过利用程序分析技术和预测模型，识别新手程序员对其程序行为的误解，（2）纠正新手程序员的误解并解释其原理，而不干扰程序员，以及（3）通过为新手程序员生成测试代码来防止未来的误解，该测试代码将捕获对程序行为的意外改变。这些工具将作为代码编辑器的扩展来实现，并将开源并公开供公众使用。此外，该项目将提供教材，帮助教员将这些工具纳入编程课程，并维护一个公开的调查期间获得的关键数据数据库。总的来说，该项目致力于通过帮助不同人群的新手程序员更好地理解他们的代码，同时学习如何编程，从而产生相当大的影响。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

The objective of this research is to investigate how programming tools can be enhanced to aid novice programmers in understanding how their program behaves. The number of people interested in learning computer programming has increased dramatically in recent years. However, a particularly challenging barrier that novice programmers face is they often hold misconceptions about how their code behaves. A key reason that novices have this issue is they often have a different definition of program correctness than professional programmers. Although current tools provide helpful feedback in fixing syntax errors, novice programmers often have considerable difficulty with semantic errors, which occur when the program compiles but behaves differently than the programmer expected. This research will investigate tool designs that identify novice programmers' misconceptions, correct the misconceptions, and prevent future misconceptions. Taking a user-centered design approach, the researcher will share findings from a series of comparative user studies and a field deployment that measure the effectiveness of the tool designs. The availability of such tools could lower the barriers that novice programmers face and improve accessibility to learn how to program.In particular, this project will develop novel tool designs that accomplish three mutually supportive goals: (1) identify misconceptions that novice programmers have about their program's behavior by leveraging program analysis techniques and predictive models, (2) correct novice programmers' misconceptions and explain the rationale without disrupting the programmers, and (3) prevent future misconceptions by generating test code for the novice programmers that will catch unintended changes to the program's behavior. The tools will be implemented as extensions to code editors, and will be made open source and publicly available for the public to use. Additionally, the project will contribute teaching materials to aid instructors in integrating the tools into programming courses, and maintain a publicly available database of key data that were obtained during the investigations. Overall, this project strives to make considerable impact by helping a diverse population of novice programmers to better understand their code while learning how to program.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.