MODEL EVALUASI DALAM MATA KULIAH DOKKAI

Abstract

Dokkai is a reading comprehension subject at the Department of Japanese Language UNJ, which is conducted from the third semester to the sixth semester. During the teaching-learning process, we have been conducting various kinds of evaluation, including formative and summative tests, and from the results we found that the students could proceed to a higher reading level. But it turns out when we also found that they are still having problems when solving the reading comprehension test in Japanese Language Proficiency Test (JLPT) which is held twice a year. So we feel the need to solve this problem by finding an effective evaluation model to measure students' reading comprehension abilities. From the results of this study are expected to make an effective evaluation model so that we can organize a more accurate evaluation of the Dokkai course. This research method is the evaluative research. Evaluative method is intended to compare the activities, events, and products with a predetermined standard. This study aims to find appropriate evaluation model to measure reading comprehension. The first step is making the same test format with the form on the JLPT test, but it seems to be ineffective to measure student learning outcomes in the Dokkai subject. So we decided to try using an interview technique. After getting the datas we found there is a difference between the results of written test and the test with interview techniques. Written test tends to provide higher score than interview results. It is because on written test, students have more time to read over again the text than in oral test. In the other hand, the student may also copy the sentence of the text that related to the question given. Or in the worst, on written test there will always a probability for student to cheat and copy the other’s answer, which they have no

chance to do such thing on oral or interview test.From all matters founded in this research, we considere that it will be better to collaborate the both tests at the time. The purpose of combining the two forms of tests is to provide more reliable datas about actual student ability on Dokkai or reading comprehension.