Lessons Learned in Designing and Implementing a Computer-Adaptive Test for English

This paper describes the lessons learned in designing and implementing a computer-adaptive test (CAT) for English. The early identification of students with weak L2 English proficiency is of critical importance in university settings that have compulsory English language course graduation requirements. The most efficient means of diagnosing the L2 English ability of incoming students is by means of a computer-based test since such evaluation can be administered quickly, automatically corrected, and the outcome known as soon as the test is completed. While the option of using a commercial CAT is available to institutions with the ability to pay substantial annual fees, or the means of passing these expenses on to their students, language instructors without these resources can only avail themselves of the advantages of CAT evaluation by creating their own tests. As is demonstrated by the E-CAT project described in this paper, this is a viable alternative even for those lacking any computer programing expertise. However, language teaching experience and testing expertise are critical to such an undertaking, which requires considerable effort and, above all, collaborative teamwork to succeed. A number of practical skills are also required. Firstly, the operation of a CAT authoring programme must be learned. Once this is done, test makers must master the art of creating a question database and assigning difficulty levels to test items. Lastly, if multimedia resources are to be exploited in a CAT, test creators need to be able to locate suitable copyright-free resources and re-edit them as needed.

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.