Analysis of Test Instrument Quality in Measuring Computational Thinking Skills in Seventh Grade Junior High School Ratio Material

Abstract

The integration of computational thinking in mathematics learning is essential to support 21st-century skills; however, assessment practices in schools still inadequately measure students’ thinking processes. This study aims to analyze the quality of a test instrument designed to measure junior high school students’ computational thinking ability on ratio material. A quantitative descriptive-evaluative method was employed involving 64 seventh-grade students. The instrument consisted of five contextual essay items developed based on computational thinking indicators: decomposition, pattern recognition, abstraction, generalization, and algorithmic thinking. Item analysis was conducted to examine validity, reliability, difficulty index, and discrimination power. The results indicate that all items were valid and the instrument showed high reliability (α = 0.81). The difficulty levels ranged from easy to moderate, while the discrimination power was categorized as sufficient to good. These findings demonstrate that the instrument is valid, reliable, and appropriate for measuring students’ computational thinking ability in mathematics learning at the junior high school level.