The virtual CAT: A tool for algorithmic thinking assessment in Swiss compulsory education

Giorgia Adorni,Alberto Piatti

2024-08-28

Abstract:In today's digital era, holding algorithmic thinking (AT) skills is crucial, not only in computer science-related fields. These abilities enable individuals to break down complex problems into more manageable steps and create a sequence of actions to solve them. To address the increasing demand for AT assessments in educational settings and the limitations of current methods, this paper introduces the virtual Cross Array Task (CAT), a digital adaptation of an unplugged assessment activity designed to evaluate algorithmic skills in Swiss compulsory education. This tool offers scalable and automated assessment, reducing human involvement and mitigating potential data collection errors. The platform features gesture-based and visual block-based programming interfaces, ensuring its usability for diverse learners, further supported by multilingual capabilities. To evaluate the virtual CAT platform, we conducted a pilot evaluation in Switzerland involving a heterogeneous group of students. The findings show the platform's usability, proficiency and suitability for assessing AT skills among students of diverse ages, development stages, and educational backgrounds, as well as the feasibility of large-scale data collection.

Human-Computer Interaction,Artificial Intelligence,Computers and Society

What problem does this paper attempt to address?

The problem that this paper attempts to solve is some limitations in the current Algorithmic Thinking (AT) assessment methods, especially in the compulsory education environment in Switzerland. Existing assessment methods, such as multiple - choice tests, programming tasks, robot activities, etc., are effective in some aspects, but they often focus on the assessment of memory and basic knowledge, lack a comprehensive evaluation of the ability to solve real - world problems, and have problems such as low efficiency and being error - prone when applied on a large scale. In addition, these methods sometimes overlook factors such as students' age differences, development stages, educational backgrounds, and social environments.

To solve these problems, this paper introduces a digital tool named Virtual Cross Array Task (CAT), aiming to overcome the limitations of existing methods and achieve large - scale AT assessment. The virtual CAT has the following characteristics:

1. **Automated assessment**: Reduce human participation through digital means and lower the risk of data collection errors.
2. **Multilingual support**: Adapt to the multilingual environment in Switzerland and support Italian, French, German, and English.
3. **Multiple interaction interfaces**: Provide programming interfaces based on gestures and visual blocks to ensure that students from different backgrounds can use it.
4. **High adaptability**: It can conduct assessments on multiple devices simultaneously, suitable for students of different ages and educational backgrounds.

Through these characteristics, the virtual CAT can not only assess students' algorithmic thinking abilities more effectively, but also provide teachers with instant feedback, helping them better adjust their teaching methods.