Algorithm design patterns: a conceptual framework for simplifying complex problems in beginner computer science education

Teaching algorithmic problem-solving remains a major challenge in beginner computer science education because novice programmers often struggle to identify appropriate solution strategies beyond programming syntax. Existing instructional approaches generally present algorithm paradigms as isolated topics, making it difficult for learners to recognise problem characteristics and transfer knowledge across different computational tasks. This study proposes the Algorithm Design Patterns (ADP) framework, a conceptual instructional model that organises five fundamental algorithm paradigms into reusable learning patterns for beginner computer science education. A conceptual research approach was adopted by synthesising literature on algorithmic thinking, programming pedagogy, instructional scaffolding, and algorithm design. The proposed framework integrates Divide and Conquer, Greedy Algorithms, Dynamic Programming, Backtracking, and Recursive Patterns into a progressive instructional structure supported by representative algorithms, classroom activities, and common learning pitfalls. By emphasising problem analysis before algorithm selection, the framework aims to strengthen algorithmic thinking and support more systematic learning. The proposed framework offers a structured pedagogical reference for introductory algorithm instruction and provides a foundation for future empirical validation in computer science education.