Predicting EFL expository writing quality with measures of lexical richness

This paper investigates the relationship between lexical richness and EFL expository writing quality and examines the predictability of lexical richness indices to EFL expository writing quality. Two hundred and seventy expository writing samples were drawn from Spoken and Written English Corpus of Chinese Learners Version 2.0. The lexical richness of the writing samples was analyzed with Lexical Complexity Analyzer, and the values of the 26 indices were calculated being the independent variables to predict the EFL expository writing quality. Besides, the writing samples were rated by three experienced raters and the average scores from the three raters were used as the dependent variable. The results of correlation analysis show that all four measures of lexical richness, i.e., lexical density, sophistication, variation, and fluency, are significantly correlated with the EFL expository writing quality, but the strength of the correlation is either low or medium. The results of regression analysis show that two indices of lexical richness, i.e., Number of Words and Noun Variation, can explain 38.5% (r = 0.620, p = 0.000) of the variance in the average score of EFL expository writing. A 10-fold cross-validation was performed and the results indicate that the model validly fits the data and can be generalized with unseen data.