Abstract
This project offers a novel, multidimensional account of how English as a Foreign Language (EFL) writing develops across the six Common European Framework of Reference (CEFR) proficiency levels (A1–C2). It profiles learner writing along four key dimensions: functional adequacy, accuracy, syntactic complexity, and lexical diversity. Addressing current gaps in learner corpus research and L2 assessment, the project tackles (1) the limited attention to functional adequacy, (2) the dominance of reductionist, mono-feature approaches to L2 writing development, and (3) the over-reliance on intuition rather than empirical evidence in the design of L2 descriptor scales for (high-stakes) assessment contexts.
The analysis draws on learner corpus data from COREFL (Corpus of English as a Foreign Language) and ICLE (International Corpus of Learner English), which will be annotated both manually and automatically for the four focal constructs. Advanced inferential methods (e.g., linear mixed-effects modeling, structural equation modeling) will be used to (1) track developmental trajectories of each construct across proficiency levels, (2) explore dynamic interrelationships among the constructs, and (3) assess to what extent functional adequacy can be predicted based on accuracy, syntactic complexity, and lexical diversity profiles. The findings will inform the development of empirically-grounded, pedagogically transparent CEFR-aligned L2 English writing descriptors.
This project will also lay the groundwork for a follow-up postdoctoral initiative aimed at developing a semi-automatic assessment tool for L2 writing. This digital interface will allow assessors to upload learner texts, annotate linguistic features (FA, accuracy, SC, LD) with standardized tags, and evaluate texts using the newly developed CEFR-aligned descriptors. To date, no such tool is openly available to learners, teachers, or researchers.
Researcher(s)
Research team(s)
Project type(s)