A Coh-Metrix-based Study On Text Features Of Expository Writings By Chinese English Learners

By means of the online text analysis tool Coh-Metrix,this research analyzes 106 text features of expository writings by English majors from five aspects(i.e.,descriptive information,lexical sophistication,syntactic complexity,coherence,and text readability)to explore the textual features that influence the quality of expository writing for Chinese English learners.With the purpose of investigating the differences in text features between the high-and the low-score writings,the correlations between text features and writing score,and the influences of text features on expository writing quality,this research adopts a combination of quantitative research and qualitative research to analyze 210 timed expository writings.Major conclusions are as follows:Firstly,the high-and the low-score writings differ significantly in 33 text features,involving five aspects of Coh-Metrix linguistic indicators.Secondly,in terms of the correlations between text features and writing score,semantic similarity of all sentences,syntactic simplicity,deep cohesion,verb cohesion etc.12 linguistic indices are significantly correlated with writing score.Lastly,according to the results of regression analysis,the four text features combined(the number of words,the semantic similarity of all sentences,the familiarity of content words,and the incidence score of temporal connectives)account for 42.4%of the variance in the evaluation of the 210 writings examined.The predictive model is:Writing score(in a hundred-mark system)=128.617+DESWC×0.121+LSAGN×23.676-WRDFAMc×0.146+CNCTemp×0.066.This study helps Chinese learners of English understand the differences in text features between the high-and the low-score writings.Meanwhile,the text features significantly related to writing quality also provide a valuable reference for the teaching and grading of expository writing.