Does Student Autonomy Make A Difference In Listening Test Performance?

This paper reports on an empirical study which set out to find out the effect of students' autonomy on test performance by comparing two different ways of assessing EFL learners' listening comprehension, an autonomous test and a traditional one. In the autonomous test, which is a tailor-made test for an autonomous listening course, the examinees are entitled to full control of the listening process within a set testing time, which means they can skip, skim, scan, pause and repeat at will while listening. The traditional test mimics a commonly-found pen-and-pencil listening test in which the examinees are deprived of the freedom to listen in their own manner as they do in the autonomous test. Both tests are computer-based and realized with a courseware specially designed for this study.One hundred and twenty second-year English majors participated in the study. Divided into two homogeneous groups, they took both tests. Each test consists of 18 multiple-choice question items on 4 pieces of news broadcast covering different topics. During the tests, the students were also required to complete two questionnaires probing into their strategy use and the playing times (the number of times the listening materials were played). Upon completion of the tests, nine students were interviewed and encouraged to reflect on how they did the tests. A teacher of the listening course was interviewed too.The main finding is that the students' performance in both tests exhibits no statistically significant difference. In other words, students' autonomy makes no difference in testing listening comprehension. The reason for the unsatisfactory performance in the autonomous test is most like inappropriate time allotment. Furthermore, the strategy use, made possible by the presence of autonomy, might not have been conducive to good performance as expected. However, the lack of better performance in the autonomous test did not undermine the test takers'enthusiasm for this type of test. Seven of the nine students who were interviewed prefer the autonomous way of testing to the traditional way.Therefore, it is recommended that when adopting an autonomous way of testing, examinees should receive training on time allotment so that the test would yield fruitful results. Due attention should be paid to the strategy use too. It is inferred that, if more emphasis was laid on the strategy of grasping the gist, the students could have excelled themselves in the autonomous test.The study is enlightening in that it offers some insights to teachers and learners engaged in this autonomous listening course as to what could lead to better performance and how to achieve it. Furthermore, it enriches our understanding of the effect of input procedures on examinees' test performance.