Teaching Computational Thinking Skills Through Scratch in K‐12 Education: A Systematic Review

ABSTRACT Background Computational thinking (CT) is increasingly critical in K‐12 education for fostering essential digital‐age problem‐solving abilities. Scratch, a widely adopted block‐based programming environment, has emerged as a promising platform for developing CT skills, yet a clear synthesis of instructional use remains limited. Objectives This paper presents a systematic literature review on instructional methods for teaching CT skills through Scratch in K‐12 settings. The review aims to: (1) identify the contexts and subject domains in which Scratch is used to teach CT; (2) explore the instructional strategies employed to integrate CT teaching through Scratch; (3) examine factors that influence CT learning outcomes; and (4) investigate challenges in implementing Scratch‐based CT instruction. Method Following PRISMA guidelines, a comprehensive search strategy was used to identify relevant literature, resulting in 46 included empirical studies. Studies were thematically coded and analysed to synthesise key findings. Results and Conclusions Scratch‐based CT instruction is common in primary and middle school settings. Most studies position Scratch as a standalone tool for CT teaching, followed by integration with science, mathematics, and STEM, with very few studies in non‐STEM subjects. Three categories of teaching approaches were identified, including task‐oriented, social‐oriented, and technology‐based approaches. Instructional effectiveness is influenced by student age, gender, and task difficulty. Key challenges include the unstructured Scratch interface, difficulties in teaching abstract CT concepts, and reliance on trial‐and‐error methods. The review highlights the need for scaffolded, developmentally appropriate teaching approaches and suggests expanding CT education into non‐STEM domains and diverse school levels.