Study On The Development Of Writing Proficiency Of English Majors In China

This study, based on the complexity theory (Larsen-Freeman,1997), with12English majors as research participants, the keystroke logging tool Inputlog as research tool, was conducted by tracing the products of computer-based writing once every two weeks over a18-week (one semester) period. It aims to reveal the nonlinearity, diversity, variability and complexity of the development of English writing proficiency.This paper attempts to answer the following four questions:(1) Compared with the first writing, does the proficiency reflected in the last writing have any progress after one semester’s training?(2) What are the respective development patterns of accuracy, fluency and complexity of the whole group over the semester?(3) What are the respective development patterns of accuracy, fluency and complexity of individuals over the semester? Do individuals have any variability?(4) What are the relationships of accuracy, fluency and complexity?The study includes the following steps:(1) Twelve participants were asked to do writing tasks on computers installed with software Inputlog at fixed time once every two weeks. They were required to do typing test for five minutes and then write an essay (about200words) within40minutes. The study collected data of nine times and altogether104essays.(2) The essays of pre-test and post-test were anonymized and numbered. Two instructors assessed the essays individually under the criteria of Jacobs’et al.(1981) ESL Composition Profile. Average score of the two raters was considered as the final writing score.(3) Accuracy of each essay was judged and calculated manually. The three computational tools, Inputlog, Web VocabProfile, and L2Syntactic Complexity Analyzer were used for the analysis of fluency, lexical complexity and syntactic complexity respectively.(4) The software SPSS18.0was used to conduct correlation analysis, paired samples t-test and partial correlation analysis.The major findings of this study are presented as follows: Firstly, all the twelve participants have fully improved in writing proficiency after one semester’s training. Paired samples t-tests suggest that the content, organization and holistic scores of participants significantly improved in the last time; accuracy and lexical complexity are significantly higher than the first time; both fluency and syntactic complexity show no significant change between the pre-test and post-test.Secondly, in terms of group averages of writing proficiency over the nine times, the development patterns of accuracy, fluency and complexity are different. The graph for each measure shows that accuracy and lexical complexity are generally on the rise with times of fluctuations; however, their fluctuations vary from each other. Fluency develops stably with fewer fluctuations, showing no remarkable progress or regress. Syntactic complexity shows no remarkable change, but seems to have a slight upward trend.Thirdly, the developmental paths of accuracy, fluency and complexity of different individuals are all not linear, and vary dramatically from one to another.Fourthly, generally speaking, fluency is significantly negative correlated with lexical complexity and syntactic complexity respectively; lexical complexity and syntactic complexity are significantly positive correlated with each other. Nevertheless, the correlations of those components of writing proficiency show noticeable differences and variability within and between individuals. Despite the differences and variability, all the participants demonstrate one thing in common:they are unable to attend to all the dimensions (accuracy, fluency and complexity) of writing proficiency at one point in time.