The Relationship Of Lexical Richness To The Quality Of ESL Learners’ Expositions

This thesis probes into the components of lexical richness and their relationship to the quality of English as a second language(ESL) learners’ expositions. The definitions of lexical richness and their significance to learners and teachers, as well as its structure and components from different perspective were reviewed from language acquisition and teaching literature. 21 measurements were selected as variables and discussed.The study was corpus-driven and a computational tool was employed to automatically calculate the 21 statistics of 3 dimensions of lexical richness, that is, lexical density, lexical sophistication and lexical variation. In the s tudy, 270 expositions from Spoken and Written English Corpus of Chinese Learners were first POS tagged and reformatted to serve as the proper input of a computational tool, Lexical Complexity Analyzer. Their relationship to each other was investigated and the correlation of lexical richness to the scores given by raters was examined.The proposition that lexical richness consists of three dimensions(lexical density, lexical sophistication and lexical variation) was demonstrated by an exploratory factor analysis, implying they are indeed three different constructs in teaching and testing. By comparing the relationship of each measurement under the three dimension to the scores given by expert raters, factors that are predictive to the quality of ESL learner s’ expositions were explored.The results of the study enlightens a deeper understanding for the concept and usage of lexical richness and its relationship to the second language teaching and testing. By exploring predictive indicators to the quality of learner’ expositions, certain measurements are that can help teachers and test designers to develop a more valid and reliable test are revealed. The study also provides evidence for the usefulness and effectiveness of a computational tool for analyzing lexical richness of ESL samples.