Author(s)

Huaqing He, Lanyu Chen

Corresponding Author

Huaqing He

Abstract

Writing assessments, especially in China, have increasingly employed the continuation task, an integrated reading-writing task in which students read an incomplete story and then finish the story logically. This paper mainly employes the Coh-Metrix text processing Software to analyze the textual features that influence the writing quality of Chinese English as a foreign language (EFL) learner in the continuation task. Based on quantitative analysis, the following results are gained. Firstly, fluency, grammatical accuracy, lexical complexity, and cohesion are correlated with the writing score to such a degree that their indices account for 15% , 31.9% , 19.4% , and 23.6% of its variance, respectively. Secondly, the predictive model of writing quality is Writing Score (in a twenty-five-mark system) = -11.844+0.017×Number of Words+2.580×Grammatical Accuracy +0.041×The Ease of Constructing Mental Images for Content Words, which accounts for 52.4% of the variance in the evaluation of continuation writing quality. The findings are beneficial for both EFL teaching and English writing assessment.

Keywords

textual features, continuation task, writing quality, EFL learners