This study aims to analyze the quality of an essay test instrument measuring students’ mathematical representation ability on eighth-grade statistics material, which was administered to ninth-grade students at SMP Negeri 1 Bandar Lampung. The analysis covers item validity, reliability, difficulty level, and discrimination index. This study employed a quantitative approach with a descriptive method. The research subjects were 27 ninth-grade students, while the object of the study consisted of six essay items on statistics that were developed based on indicators of mathematical representation ability. The results show that only three items (numbers 2, 4, and 6) meet the validity criteria, while items 1, 3, and 5 are invalid. The reliability coefficient of the instrument is 0.685, which falls into the moderate category and is therefore acceptable. All items are categorized as easy based on the difficulty index analysis. However, the discrimination index analysis indicates that all items have low discrimination power and thus fail to distinguish between high- and low-ability students. These findings imply that the instrument requires comprehensive revision, particularly in terms of difficulty level and item quality, in order to more accurately measure students’ mathematical representation ability in statistics.
Copyrights © 2026