摘要：The purpose of this study is to describe the use of Classical Test Theory (CTT) to investigatethe quality of test items in measuring students' English competence. This studyadopts a research method with a mixed methods approach. The results show that mostitems are within acceptable range of both indexes, with the exception of items in synonyms.Items that focus on vocabulary are more challenging. What is surprising is that theshort answer items have an excellent item difficulty level and item discrimination index.General results from data analysis of items also support the hypothesis that items thathave an ideal item difficulty value between 0.4 and 0.6 will have the same ideal item discriminationvalue. This paper reports part of a larger study on the quality of individual testitems and overall tests.
