Abstract
Providing immediate computerized feedback to improve students’ writing skill at individualized pace has been a new technological approach developed for English learning. Recently, new automatic grading system which claims to provide diagnostic feedback on all aspects such as organization, word usage, and grammar of an essay is available now. A preliminary observation conducted in the Spring semester of 2003 found that students did benefit from computerized feedback on grammar and organization, but they could not obtain from such system suggestions for word usage, an area in which most EFL students are in great need for help. To meet students’ needs for word usage, the current study introduced to students, in addition to an automatic essay grading system, a bilingual concordancer which is advocated as beneficial for word usage in an EFL context (Lin, 2003). The purpose of this study was thus two-fold in investigating how effective (1) a computerized feedback system and (2) a bilingual concordancer were for college EFL students’ writing, revision, and self-correction of errors.Nineteen freshmen enrolled in a writing class in a university in northern Taiwan participated in this study in Fall 2003 and were asked to compose three writings along the research process. Two online learning tools were applied to help enhance students’ writing skill--an automatic grading system, MyAccess (http://www.gomyaccess.com) for both grading student writing and providing feedback, and a web-based bilingual concodancer, TOTALrecall (http://candle.cs.nthu.edut.tw/candle) for helping students with word usage problems in their drafts or self-correcting errors highlighted by the teacher. Instruments used were two questionnaires (background and attitude); checklists on which students recorded feedback used from MyAccess and words looked up from TOTALrecall.The analysis of students’ holistic scores indicated that students’ writing was significantly improved with the help of the feedback report from MyAccess. Also,, students considered MyAccess helpful in providing feedback for revision according to the responses from the evaluation questionnaire As for the effectiveness of TOTALrecall, over half of students’ errors were corrected with the tool, except for the first writing which might result from the instability of the tool and students’ unfamiliarity. In addition, TOTALrecall was found to be most helpful in providing information on word usages for nouns and less so in adjectives. Moreover, as shown in the evaluation questionnaire, students also regarded TOTALrecall useful for providing information on word usage and collocation. They also indicated that the parallel corpus made the learning of English words and meanings easier.. Three pedagogical implications are drown from the current study. First, there should be time allotted for learners’ computer literacy and familiarity with the target tools before using tools for computer assisted learning. Second, instructors should teach the students to make best use of the tools out of classroom so that the limited class hours could be effectively used. Third, bilingual concordancing exercises related to collocations and synonyms could be used to expend and deepen students’ vocabulary capacity. Future research could continue exploring students’ errors on word usage and enlarging error data based on the findings of the current study. Online error exercises are in need to help students improve their word use in writing. To conclude, it is hoped that the encouraging results from the current study could inspire researchers to continue develop CALL software and at the same time, invite more efforts devoted to the field of CALL.