Evaluation is a data activity used to measure the degree to which a goal is achieved. The teacher evaluates the students using a test because the test is a tool that can help the teacher measure the students' ability or competence. A test can be a good test if it fulfils several requirements such as validity, reliability, difficulty level, discrimination power, and distractor effectiveness. One of the efforts to know the quality of a test is by analyzing test items.

In this study, to know the quality of a test, the researcher used descriptive qualitative and quantitative approaches to present and describe validity, reliability, difficulty level, discrimination power, and distractor effectiveness. The research object was the English summative test of eighth-grade students made by MGMP in SMPN 26 Banjarmasin. The researcher collected the data using documents such as students’ answer sheets, English summative test, syllabus, and answer key.

This research showed that the English summative test lacked on content validity because the test items did not measure listening, speaking, and writing skill. Construct validity on the test items was also low because test items just measured reading, grammar and vocabulary skills. The reliability was in the high category (0,77). The level of difficulty was in the moderate category (46%). The discrimination power was in the poor category (54%), and it is suggested that the teacher needed to revise. Lastly, the effectiveness of distractors was in a good category (41%).

Based on the research findings, English has four essential skills to be measured to show the students' ability. The English summative test only measured reading, grammar, and vocabulary without measuring listening, speaking, and writing. It is also suggested that the test maker or teacher should reconsider providing listening, speaking, and writing to be included in the test to stimulate the ability of students to face the next English test in the next level. Besides, the test maker or teacher also gives more attention to the quality of test items so that the test can have higher quality and the learning outcome of the test can function well.