Scholarlink

Background <p>In the modern digital era, computational thinking (CT) is a fundamental skill that has been influenced by advances in artificial intelligence (AI). While there are various resources for teaching and evaluating CT, the integration of AI principles into CT education remains scarce. In this study, we proposed the Computational Thinking in AI Training Test (CTAT), tailored for higher education students in computer sciences (CS) and AI-related programs. The test was developed using the Evidence-Centered Design (ECD) framework, resulting in 34 multiple-choice items being created. Validation was conducted through expert review and cognitive interviews, with refinements made based on feedback. A field test was then conducted with 461 higher education students (Year 1–2) from two vocational schools and two colleges in southern China. The data were analyzed using the Item Response Theory (IRT) approach, along with other psychometric tests.</p> Results <p>Results indicate that CTAT exhibited robust psychometric properties, suggesting its appropriateness for the target student population. Analysis of test performance revealed that students’ CT skills improved as they progressed through their educational pathways, with no significant gender-based differences observed. Students’ response patterns to the items reveal that they tend to have difficulties in identifying appropriate data representation, applying logical operators in a correct sequence, and differentiating loop structures.</p> Conclusions <p>The study provides a valid and reliable tool for assessing CT within an AI training context at the higher education level, and findings on students’ challenges in solving CT problems provide practical insights for CT and AI tertiary education.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrating AI into computational thinking: development and validation of an assessment tool for higher education students

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月11日

查看全文
  • Shile Zhang,
  • Shuhan Zhang

摘要

Background

In the modern digital era, computational thinking (CT) is a fundamental skill that has been influenced by advances in artificial intelligence (AI). While there are various resources for teaching and evaluating CT, the integration of AI principles into CT education remains scarce. In this study, we proposed the Computational Thinking in AI Training Test (CTAT), tailored for higher education students in computer sciences (CS) and AI-related programs. The test was developed using the Evidence-Centered Design (ECD) framework, resulting in 34 multiple-choice items being created. Validation was conducted through expert review and cognitive interviews, with refinements made based on feedback. A field test was then conducted with 461 higher education students (Year 1–2) from two vocational schools and two colleges in southern China. The data were analyzed using the Item Response Theory (IRT) approach, along with other psychometric tests.

Results

Results indicate that CTAT exhibited robust psychometric properties, suggesting its appropriateness for the target student population. Analysis of test performance revealed that students’ CT skills improved as they progressed through their educational pathways, with no significant gender-based differences observed. Students’ response patterns to the items reveal that they tend to have difficulties in identifying appropriate data representation, applying logical operators in a correct sequence, and differentiating loop structures.

Conclusions

The study provides a valid and reliable tool for assessing CT within an AI training context at the higher education level, and findings on students’ challenges in solving CT problems provide practical insights for CT and AI tertiary education.