Using TAASSC to Investigate Fine-Grained Grammatical Complexity in Reading Texts of Two High-Stakes English Tests in China

Abstract

Fine-grained grammatical complexity measures are better predictors than large-grained indices of register variation and writing of different levels in that they provide in-depth explanation of what accounts for syntactic complexity. Computer technologies have made it possible to study these fine-grained measures based on large corpora. Using the Tool for Automatic Analysis of Syntactic Sophistication and Complexity (TAASSC) (Kyle, 2016), the present study investigates phrasal and clausal complexity in reading texts of two high-stakes English tests in China (CET-4 and NETEM) different in difficulty level. The results showed that clausal and phrasal complexity indices could predict the two tests. More importantly, in combined analysis of the two types of measures, the results demonstrated that noun phrases, particularly those with phrasal modifiers, were better predictors of the more difficult test NETEM. These findings support the importance of noun phrases and the diverse modifiers they take in informational texts. Findings of this research will help college students advance their English learning with a specific purpose and in a right direction.