Assessing Computational Thinking: The Relation of Different Assessment Instruments and Learning Tools.

The relevance of computational thinking as a skill for today's learners is no longer in question, but every skill needs an assessment system. In this study, we analyze two validated instruments for assessing computational thinking - the CTt (Computational Thinking Test) and the CTS (Computational Thinking Scale). The study involved 49 students in grades 8 and 9 (age 14-16). Prior to the study, students in both grades were taught computational thinking differently. One group learned computational thinking by completing tasks and creating projects in Scratch, the other group learned by completing tasks in "Minecraft: Education Edition". The students were asked to take the CTt and CTS tests. The nature of these tests is different, one is computational thinking diagnostic tool, the other is a psychometric self-assessment test consisting of core abilities (subconstructs) important for computational thinking. The aim of this study was to determine how these tests related to each other and whether students' gender and the different tools chosen to teach computational thinking had an impact on the level of computational thinking knowledge and abilities acquired based on the tests. The results have shown that the scores of the two tests correlated with each other only for male students' subgroup. For a whole group CTt scores correlated only with CTS algorithmic thinking subconstruct. The results have also shown that teaching tools do have an impact on the acquisition of different computational thinking concepts skills: students taught with different tools had different test results. This study provides useful implications on computational thinking teaching improvement and its assessment better understanding.