计算思维起源于计算机科学，是一种能够让任何领域学生受益的技能。在众多已经被证实能培养学生计算思维的方式中，编程训练是最常用的一种方式。而在编程训练的众多方式中，配对编程已经被证实比独立编程更加有利于学生计算思维的培养。与此同时，算法和调试是计算思维活动中至关重要的两个部分，了解学生在算法和调试过程中的策略发展有助于教师更好地引导学生学习编程。另外，以往针对配对编程的研究大多采用实验研究法，这种研究方法关注编程学习的绩效提升，而忽视了配对编程中，学生编程策略和行为模式的发展变化。所以本研究将算法、调试和认知策略发展联系起来，通过微观发生法研究，在配对编程中，小学生在计算思维培养中的算法和调试策略是如何发展的。
本研究选取Apple Swift Playgrounds软件作为编程工具，软件中的《学习编程1》课程活动中的三个主题活动中的关卡作为教学内容，小学四至六年级学生作为研究对象，通过游戏化教学的方式在为期3课时，每课时40分钟的活动中进行本次研究。本研究通过课堂观察法和访谈法收集了活动中学生的活动录像、访谈记录，以配对组为单位分析学生在三个主题的四个阶段中的算法策略和调试策略的发展变化情况。
经过数据分析，本研究得到以下研究结论：
（1）命令语句主题活动中主要的算法策略是最短路线策略。在这一主题活动中，算法策略整体呈现出从低级到高级不断发展的趋势，并且得到了最大程度的迁移。
（2）命令语句主题活动中主要的调试策略是不断试错策略。在这一主题活动中，调试策略整体呈现出从低级到高级不断发展的趋势，并且得到了部分程度的迁移。
（3）函数主题活动中主要的算法策略是创建简单函数策略。在这一主题活动中，算法策略整体呈现出从低级到高级不断发展的趋势，并且得到了部分程度的迁移。
（4）函数主题活动中主要的调试策略是调试器策略。在这一主题活动中，调试策略整体呈现出从低级到高级不断发展的趋势，并且得到了部分程度的迁移。
（5）For循环主题活动中主要的算法策略是循环策略。在这一主题活动中，算法策略整体以高级为主，并且没得到迁移。
（6）For循环主题活动中主要的调试策略是深度遍历策略。在这一主题活动中，调试策略整体以高级为主，并且得到了部分程度的迁移。
最后，本研究希望未来有更多研究关注配对编程中K12学生的策略发展。通过对K12学生在配对编程活动中的策略发展变化分析，进一步改进现有的K12学生的编程教材、教学策略，提高学生的计算思维能力。

Computational thinking has its roots in computer science and is a skill that can benefit students in any field. Among the many ways proven to develop computational thinking in students, programming training is one of the most commonly used. Among the many ways of programming training, pair programming is more beneficial than stand-alone programming for developing computational thinking. At the same time, understanding students' strategy development during algorithms and debugging, two crucial parts of computational thinking activities can help teachers better guide students to learn to program. In addition, most previous studies on pair programming have used experimental research methods, which focus on the performance improvement of programming learning and ignore the developmental changes of students' programming strategies and behavioral patterns during pair programming. Therefore, this study links algorithmic, debugging, and cognitive strategy development to investigate, through a micro-occurrence approach, how elementary school students' algorithmic and debugging strategies develop in the development of computational thinking in paired programming.
Apple Swift Playgrounds application was selected as the programming tool, levels from the three thematic activities in the Learning to Program 1 curriculum activities in the software were used as the instructional content, and elementary school students in grades four through six were used as the subjects of this study, which was conducted through a gamified instructional approach in a three-class period of 40 minutes each. This study collected video recordings of students' activities and interview transcripts through classroom observation and interview methods to analyze the changes in students' algorithmic and debugging strategies developed during the four stages of the three themes in paired groups.
After data analysis, the following research findings were obtained from this study: (1) The primary algorithmic strategy in the command statement thematic activity was the shortest route strategy. In this thematic activity, the algorithmic strategy showed a trend of continuous development from low to high levels, and it was migrated to the maximum extent. (2) The primary debugging strategy in the command statement thematic activity is the trial-and-error strategy. In this thematic activity, the debugging strategy tends to evolve from a low to a high level and is partially migrated. (3) The primary algorithmic strategy in the function theme activity is the create a simple function strategy. In this thematic activity, the overall trend of algorithmic strategies evolves from low to high level and is partially migrated. (4) The primary debugging strategy in the function theme activity is the debugger strategy. In this thematic activity, the debugging strategy as a whole tends to evolve from low-level to high-level and is partially migrated. (5) The primary algorithmic strategy in the For loop thematic activity is the loop strategy. In this thematic activity, the algorithmic strategies are mainly high-level and have not been migrated. (6) The deep traversal strategy is the primary debugging strategy in the For loop thematic activity. The debugging strategy was mainly advanced and partially migrated in this thematic activity.
Finally, this study hopes that more future research will focus on K12 students' strategy development in paired programming. By analyzing the changes in K12 students' strategy development in paired programming activities, we can further improve the existing programming materials and teaching strategies for K12 students and their computational thinking skills.