Measuring computational thinking – new tools for schools
Computational thinking is widely regarded as a key competence in the digital age. A research team has developed age-appropriate tasks and tools to measure this skill in school environments.
Computational thinking is the ability to create computer-assisted problem-solving methodologies. Students should learn to break down problems into smaller components, devise effective solution strategies and present those solutions in a way that both humans and computers can understand and execute.
Experts agree that computational thinking is a key competence in the digital age, which is why it features in both national and cantonal education policies. But how can we determine whether these support measures are actually effective? Until now, Switzerland has lacked suitable tools to systematically measure the development of computational thinking skills among schoolchildren.
As part of NRP 77, the research team led by Francesco Mondada has developed new tools and protocols, which are easy for teachers to use and enable semi-automated assessment on a broad scale in compulsory education.
Key findings
The project demonstrated that even very young children – regardless of gender – can solve complex algorithmic tasks. With specially designed tasks, computational thinking skills can be reliably measured across all levels of compulsory education. The developed tools provide scalable, automated and age-independent assessment of these skills, without the need for complex technologies such as AI or tutorial systems.
“It was surprising to observe that some very young children of pre-school age were able to develop extremely complex algorithms,” noted Mondada, project leader and professor of artificial intelligence and robotics at EPFL.
Relevance in policy and practice
For teachers: All tasks, tools and protocols developed are freely available online in German, French, Italian and English. They can be used for both formative assessment and as classroom learning resources from kindergarten through secondary school. The tasks are age-appropriate, require no complex technical systems and provide schools with a practical, scientifically reliable tool for the targeted promotion and assessment of computational thinking skills.
For education policy: The developed tools enable the large-scale, standardised assessment of computational thinking skills. This supports the empirical evaluation of cantonal or national support measures and informs the development of evidence-based education policies.
Available via open-source software (The virtual Cross Array Task (CAT) application)
Three main messages:
- From a very young age, schoolchildren of both genders are able to develop complex algorithms and generally demonstrate computational thinking skills.
- Well-designed computational thinking tasks allow the same activities to be used to assess and support these skills throughout a child’s compulsory schooling and beyond.
- Comprehensively assessing children’s computational thinking skills does not require complex tasks or methods. In particular, neither artificial intelligence nor tutorial systems are needed to conduct a reliable, scalable and automatable assessment.
To see more NRP 77 projects regarding ‘digital transformation’, click here: