ABSTRACT

Purpose: This study aims to describe the characteristics of English Final Examination items at SMA Negeri 5 Malang using the item response theory approach. Methodology: This research is quantitative descriptive research with a sample size of 344 students of class XII SMA Negeri 5 Malang. The data was taken from the results of the English final test in the form of multiple-choice questions with 5 choices, totaling 40 items. Findings: The results showed that the English final examination test instrument (1) proved to be valid, as indicated by 40 items having a loading factor> 0.3; (2) proved to be reliable as indicated by the reliability coefficient> 0.794; (3) the analysis of the level of difficulty shows that 32 items (80%) have a good level of difficulty category so that they can be used in the next assessment, while the 8 items (20%) do not meet the criteria for a good level of difficulty so they need to be revised or eliminated; and (4) the analysis of differentiating power shows an average index of 0.518 and the index of differentiating power of 40 items (100%) is in the range of 0 to +2. This research is limited to item analysis with a limited sample. Implications for Research and Practice: The implications of the study lie in the fact that a wider sample can be taken to be able to analyze up to 3PL. Moreover, the items are only focused on multiple choice, they have not explored the analysis of the quality of the descriptive items which are actually a widely chosen option in student language assessment. Future research is expected to be able to analyze English test items in the form of a descriptive test on a wide scale.