Keywords

L2 writing proficiency assessment, lexical diversity measurement, timed writing, English for academic purposes, English as a foreign language, second language writing

Abstract

Several researchers (e.g., Gebril & Plakans, 2016; Crossley & McNamara, 2012) have noted a relationship between lexical diversity (LD) and writing proficiency. Previous research has indicated that typical LD measures can account for as much as 44 % of variability in writing proficiency (Crossley et al., 2011). However, it remains unclear which of the numerous LD measures (or combination of measures) that exist best predict writing proficiency (e.g., McCarthy & Jarvis, 2010). The present study addresses these issues using scored writing assessment samples from 911 ESL students. LD was assessed with a battery of ten measures through the Python lexical-diversity package (Kyle, 2020). A PCA was conducted to determine how these LD measures varied in deriving LD scores. Writing proficiency scores were then predicted by using linear mixed-effects regressions that included either a single LD measure or multiple measures. The results indicated that LD is a meaningful predictor in timed writing proficiency assessments, multiple LD measures have more predictive power than individual measures, timing has a slight effect on LD, and some LD measures are highly correlated while others are not. Implications for automated writing proficiency assessment and rubric design as well as LD measurement research are explained.

Original Publication Citation

Woods, K., Hashimoto, B., & Brown, E. K. (2023). A multi-measure approach for lexical diversity in writing assessments: Considerations in measurement and timing. Assessing Writing, 55, 100688.

Document Type

Peer-Reviewed Article

Publication Date

2023

Publisher

Assessing Writing

Language

English

College

Humanities

Department

Linguistics

University Standing at Time of Publication

Assistant Professor

Download

Included in

Linguistics Commons

Share

COinS