Predicting lexico-grammatical competence in L2 Spanish writing through multiple lexical diversity measures

Abstract The present study compares several lexical diversity (LD) measures to determine which measure, or measures, best predict receptive lexico-grammatical ability in written L2 Spanish, and whether a composite LD score is better than any single measure. We analyzed 1,225 written responses with eight different LD measures: six popular measures, a composite score based on Principal Component Analysis (PCA) of those six measures, and the traditional TTR to be used as a baseline against which to compare the other measures. Other predictor variables included the age and gender of the writers, the age of first exposure to Spanish, the number of years studying Spanish, and study abroad participation. The results of a series of mixed-effect logistic regression models suggest that the composite LD measure is best, and that among the LD measures studied, two predicted lexico-grammatical ability nearly equally well. We conclude with the recommendation that L2 language researchers use multiple LD measures, including a composite measure based on PCA, rather than any single LD measure.