What linguistic features distinguish and predict L2 writing quality? A study of examination scripts written by adolescent Chinese learners of English in Hong Kong

Research output: Contribution to journalArticlepeer-review

29 Citations (Scopus)

Abstract

While such linguistic features as lexical complexity and syntactic complexity have been used to analyse the texts written by college-level L2 learners and experts in previous research, the findings may not be applicable to the writing of secondary students in their learning contexts. This study, based on a corpus of 64,060 words comprising 120 high-, mid- and low-scored reports and letters written by candidates of the HKDSE English Language Writing paper, adds to the existing literature by investigating how lexical and syntactic complexity features can distinguish and predict the quality of L2 writing. Automated tools were adopted to measure the lexical and syntactic complexity features of the writings, followed by the use of one-way ANOVA and post-hoc analyses to examine the differences among the three levels of examination scripts. Correlation and regression models were also performed to test the predictive power of lexical and syntactic complexity features. It was found that lexical complexity, when compared to syntactic complexity, is a stronger predictor of L2 writing quality. The impact of text types, examination prompts and learners’ English proficiency on L2 writing quality are discussed, and the implications for the teaching and assessment of L2 writing are considered.

Original languageEnglish
Article number102461
JournalSystem
Volume97
DOIs
Publication statusPublished - Apr 2021

Keywords

  • L2 writing quality
  • Letter
  • Lexical complexity
  • Report
  • Secondary Chinese English learners
  • Syntactic complexity

Fingerprint

Dive into the research topics of 'What linguistic features distinguish and predict L2 writing quality? A study of examination scripts written by adolescent Chinese learners of English in Hong Kong'. Together they form a unique fingerprint.

Cite this