THE EXPLOITATION OF E-WRITING IN AN EFL CLASSROOM: POTENTIAL AND CHALLENGES

• Main Authors: Su-Hwa LinHuang, 林黃淑化
• Other Authors: Ching Chung Guey
• Format: Others
• Language: en_US
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/31486344540816976185

碩士 === 義守大學 === 應用英語學系碩士班 === 98 === Technology has had a big influence on foreign language education, particularly since films, computers and the Internet began to be used by both teachers and learners. Learning writing skills through automated writing evaluation (AWE) is one example of this trend. AWE is claimed by its designers to assist teachers and students in the teaching and learning of writing skills. An important pedagogical question is whether or not AWE is actually effective in improving student writing skills. A California survey by Warschauer and Grimes (2008) showed that AWE programs do seem to save some teacher time on marking essays but that student use of AWE resulted in their making only minor revisions and that further surveys were needed in other schools. The purpose of this study is to explore whether the use of an automated writing evaluation program with students accumulating an e-writing portfolio is beneficial to Taiwanese university students’ learning to write in an autonomous learning context. The AWE program MY Access was specially installed for the use of students in one class at a southern Taiwan university. This was an experiment to observe whether fifty-eight 4th year English language major students using the program independently to write essays on four different topics (two narrative, two persuasive) over a three-month period was helpful in their learning writing skills. A total of 200 essays were submitted. Records of student submissions revealed that 88% of the students participated in the program by submitting at least one essay with 53.3% submitting more than one essay, most of whom did one essay and then moved on to the next topic. This pattern of engagement highlighted the differences between autonomous learning and learning which is teacher-directed. An attitude questionnaire designed to investigate student use, autonomy and motivation was administered at the end of the period to 36 students and in-depth interviews were also conducted with eight students by random sampling. Students generally felt the program helped them learn to write better, for example, by making them more sensitive to spelling errors, although no evidence was found to show that student essay scores improved over a small number of essay revisions. A degree of correlation (0.539) was observed between student motivation and the number of essays they submitted. In-depth interviews revealed that students felt the advantages of using the program outweighed the disadvantages. Comparing the error analyses of Microsoft Word, MY Access and human raters resulted in MS Word identifying a mean error rate of 6.3 errors per essay compared to 13.2 found by MY Access on the same essays compared with 18 errors being found by human raters. Scoring of essays by the program, particularly at the MY Access lower proficiency levels of 1, 2 and 3 correlated well (0.811) with judgments of human raters although at higher writing proficiency levels of 4 and 5 some discrepancies were observed with human raters sometimes seeing more value in the content of an essay and overlooking minor errors. Results of the study suggest that an experiment teaching writing to one group by traditional means and to another matched group using automated writing software and comparing outcomes would be a logical extension of the present study. Such a study could be conducted over say a one-semester or a one- year period to give a fair trial.