A strategy to assess computational thinking
Antonio González-Torres, Lilliana Sancho-Chavarría, Magaly Zúñiga-Céspedes, Jorge Monge-Fallas, José Navas-Sú · 2021 International Conference on Computational Science and Computational Intelligence (CSCI) · 2021
Computational Thinking (CT) is a method of analysis and problem solving that enables the development of life skills and positively influences the social and competitive conditions of people. This type of thinking promotes learning through the application of planning, analysis, abstraction and decomposition of problems into their constituent parts. In addition, PC contributes to the formation of critical thinking, creativity and cooperation between people. In education, the mechanisms used to foster CP are diverse; among them, learning through programming. Databases of source code generated in learning exercises using the CP method are a valuable tool for discovering learning patterns. However, there are no robust and scalable methods for studying the huge source code bases of learning products to recognize the predominant patterns and determine the effectiveness of the method. Consequently, a comprehensive analysis of learners’ skills, knowledge, outcomes and performance levels is necessary to identify points of improvement to meet learning objectives and modify curricula and exercises. This paper presents the joint research between the Costa Rica Institute of Technology (TEC) and the PRONIE MEP-FOD program of the Omar Dengo Foundation. Its objective is to automatically analyze the source code created by the students, evaluate the Computational Thinking skills they developed and compare the results with the learning objectives established in the study plans.