A Digital Game-Based Model for Assessing Computational Thinking Skills

Abstract

Computational thinking (CT)., considered to be one of the three major scientific thinking styles of human beings., is a competency fundamental to computing education and beyond. However., as a cross-domain competency., appropriate assessment design and method remain equivocal. Indeed., the majority of the existing assessments have a predominant focus on using scales or test to measure programming proficiency and neglecting process evaluation in which CT can also be manifested. Game-based learning assessments (GBLA) are able to assess tacit knowledge by avoiding jargon in test items., constructing irrelevant material., and reducing anxiety., avoiding the barriers of traditional assessments in assessing students' abilities. To broaden the promotion and practice of CT., it is necessary to integrate diverse problem-solving contexts using a digital game-based assessment model to measure CT. We invited 13 experts to participate in three rounds of the Delphi method and constructed a CT skills assessment model based on the digital game. The model had good consistency and was divided into four primary indicators., namely “Analysis”., “Design”., “Iteration”., and “Generalization”., which were weighted as 0.26., 0.11., 0.24., and 0.39., respectively and 14 secondary indicators. Taken together., the study demonstrated the feasibility to expand from traditional assessment methods to integrating digital games in measuring CT skills in a comprehensive manner.