As computational thinking (CT) gains more attention in K-16 education, problem-solving has been more emphasized as a core competency that can be found across various domains. To develop an evaluation framework that reveals students’ problem-solving competency, this study examined solutions for the Bebras Computing Challenge which requires students to utilize problem-solving skills in a CT domain. A total of 246 solutions of three Bebras tasks were analyzed based on a qualitative content analysis method and four levels of solutions were identified. The solution levels revealed how students (1) failed to understand a problem (No solution), (2) solved the problem but failed to identify the pattern (Premature level), (3) identified principles embedded in the problem but failed to apply them to devise an automized solution (Intermediate level), and (4) identified principles and solved the problem by applying them (Advanced level). This study presented solution levels across Bebras tasks and discussed how task difficulty affected student solutions differently. Implications for teaching problem-solving skills were discussed.

随着计算机思维（CT）在K-16教育中得到越来越多的关注，解决问题作为一种核心能力而得到了越来越多的重视，这种能力可以在各个领域找到。为了开发一种评估框架来揭示学生的解决问题的能力，本研究研究了Bebras计算挑战的解决方案，该挑战要求学生在CT领域中利用解决问题的技能。基于定性内容分析方法，共分析了三个Bebras任务的246个解决方案，并确定了四个级别的解决方案。解决方案级别揭示了学生（1）如何无法理解问题（没有解决方案），（2）解决了问题但未能识别出模式（过早解决方案），（3）识别了问题中嵌入的原理但未能应用它们设计一个自动化的解决方案（中级），（4）确定原理并通过应用这些原理解决了问题（高级）。这项研究提出了Bebras任务的解决方案级别，并讨论了任务难度如何不同地影响学生的解决方案。讨论了教学解决问题的能力。