Kesulitan Berpikir Komputasional Siswa dalam Menyelesaikan Masalah Transformasi Geometri setelah Pembelajaran Berbasis Masalah Berbantuan GeoGebra

Ruth Helen Simarmata (Mathematics Education, Universitas Sriwijaya, Palembang, Indonesia)
Esterlina (Mathematics Education, Universitas Pendidikan Indonesia, Bandung, Indonesia)
Fifi Dhafia Az Zahra (Mathematics Education, Universitas Sriwijaya, Palembang, Indonesia)

Inovasi Matematika (Inomatika)
Vol. 8 No. 2 (2026): Inovasi Matematika

Computational thinking is an important mathematical process in problem solving that encompasses the abilities to analyze situations, develop strategies, recognize patterns, and design systematic procedures. Understanding students’ computational thinking cannot be derived solely from overall scores but also requires an explanation of where their reasoning processes break down. This study aimed to diagnose students’ computational thinking difficulties, analyze how these difficulties interacted and propagated throughout the problem-solving process, and identify the reasoning mechanisms underlying the identified difficulty pathways after GeoGebra-assisted problem-based learning. This descriptive study employed quantitative and qualitative approaches and involved 34 Grade 11 students from a public senior high school in Palembang, Indonesia. Data were collected through three computational thinking test items, students’ written responses, classroom observations, and interviews. The test assessed four indicators: decomposition, pattern recognition, abstraction, and algorithmic thinking. Overall, 19 students, or 55.88%, were in the moderate category. Pattern recognition showed the highest achievement at 62.74%, followed by decomposition at 50.32% and abstraction at 49.34%, while algorithmic thinking was the lowest at 44.44%. Students’ difficulties included incomplete organization of information, failure to connect transformation types with appropriate matrices, inability to select relevant information, errors in operations, and weak result verification. The findings indicate that computational thinking indicators do not operate independently but function as interdependent processes, with matrix representation serving as a critical bridge between conceptual recognition and procedural execution. Practically, the findings provide a basis for diagnostic assessment and targeted remediation according to the source of difficulty, including problem structuring, matrix representation, procedural execution, and verification. This study contributes an indicator-based diagnosis of students’ computational thinking difficulties.