A Quantitative Study Of Language Development In EFL Students' Writing At A University In China

Skehan (1992, 1996a) distinguished between three dimensions of language development: complexity, accuracy and fluency, and argued achieving an effective compromise between these dimensions leads not only to the effective communication but also to long-term linguistic development. This study belongs to the developmental index study which attempts to gauge the development of learners at known proficiency levels with complexity, accuracy and fluency measures. Though there have been a large amount of developmental index studies abroad since 1970s, it is rarely seen in China. This study aims to find how and at what rate the language develop can be achieved by undergraduate students and find the concrete ways that language development can be demonstrated in their writing. The specific research questions are: (1) what are the developmental trends of syntactic complexity, fluency and accuracy; (2) what are the relationship among syntactic complexity, fluency and accuracy. The developmental measures used in this study are mainly employed from Wolfe-Quintero's (1998) best measures in their synthesis research.A cross-sectional study was conducted with the participants of four groups of EFL students from four semesters at a Chinese university, with 30 students in each group. Altogether 120 argumentative essays were collected under the examination condition at the end of 2007 Autumn Semester and 2008 Spring Semester and then submitted to quantitative analysis utilizing the objective measures concerning the three dimensions of language proficiency: complexity as assessed by sophistication measures including the sentence coordination ratio, subordinate ratios, reduced clause ratio, and variation measures in terms of types of subordinate clauses; fluency by the mean length of sentences, T-units, error-free T-units and clauses; and accuracy by the ratio of errors per T-unit, error-free T-unit ratio as well as specific classification of syntactic errors. One-way ANOVA and/or non-parametric tests were performed to examine whether any differences were significant and testify the distribution of the differences among them. Correlation analysis was also performed to explore the inter-relationship between complexity, fluency and accuracy measures.Results indicate that across the four groups of students there was a trend of development in all the three dimensions of syntax, however, the paths were inconsistent, and the three dimensions interacted in supportive and competitive way. The trade-offs were discovered between complexity/fluency and accuracy when the second group improved syntactic complexity mainly by subordination, however, made most syntactic errors in writing. Following a semester of backsliding in both syntactic complexity and fluency, the fourth group achieved a more balanced development in all the three dimensions. Subordinate clauses were reduced to phrases and fewer errors occurred in their compositions. The correlation analysis shows that the fluency measures and complexity measures are closely related, which explains why the development in fluency shares the similar trend with that of complexity.The findings demonstrated the developmental features of the students'L2 language. The suggestion is that the learners and teachers should attach more importance to the improvement of syntactic complexity and try to achieve a balanced development in complexity, fluency and accuracy. The methodology of research should be further pursued in language assessment, language testing, and teaching.