Test items analysis is essential to be done since it can improve the quality of the test. A quality test can be used effectively to measure students’ achievement in a certain domain. A test administered to students should fulfill some criteria of good test, such as the validity, reliability, difficulty level and discrimination power.
This study is aimed to analyze the English Mid-Term test items at the first semester of the eight grade students of MTs. Mathalibul Huda Mlonggo in the academic year of 2016/2017 which was constructed by the English teacher. The researcher analyzed it based on the criteria of good test included the face validity, content validity, reliability, difficulty level and discrimination power. The researcher described the result of each of the analysis and concluded the quality of the test items. This study was designed as descriptive analysis research using qualitative approach. The writer only used documentary study in collecting the data.
From the results of analysis of 25 items analyzed, the test items had bad face validity. The writer found 92% or 23 (twenty three) items were error. This means, there were only 8% or 2 (two) items categorized as appropriate items. This means that the items need to be revised. On the other hand, the test items had good content validity. This considered from the research finding that there were 88% or 22 (twenty two) items related to the basic competence and indicator. The reliability analysis showed that the items reached r11 = 0.691. This means the test had high reliability. From difficulty level analysis, the test were categorized into good test items because the writer found 18 (eighteen) items (72%) were categorized as ideal items or having P around 0.62. There were 2 (two) items or 8% were founded as very easy items (P above 0.90). The rest 5 (five) items or 20% were categorized as very difficult items (P below 0.20). The test is called ideal items because most items found as ideal items which are not too easy and too difficult. In discrimination power term, the test items had poor discrimination power. There were 20 (twenty) items (80%) were poor items category or having D 0.00 – 0.20 and 5 (five) items (20%) were satisfactory since they reached D 0.20 – 0.40.
Finally, the researcher concluded that the English Mid-Term test items at the first semester of the eight grade students of MTs. Mathalibul Huda Mlonggo in the academic year of 2016/2017 was categorized as a poor test items. The test items should be revised for the better measurement of students’ achievement.