A text grammar-based approach for text complexity measurement across cefr levels[tesis]

Abstract

This research aims to present a novel approach to text complexity and readability measurement. The study tested 4 first-year undergraduate students, divided into 4 groups based on their linguistic performance according to the Common European Framework of Reference, and 9 Teachers of English as a Foreign Language who meet the characteristics to be labelled as experts. The data was collected through a series of exercises based on an adapted variable-complexity text, such as grammaticality judgements and comprehension questions, plus a retrospective report which allowed to collect quantitative and qualitative information. Based on the results obtained, this approach to text complexity is partially validated, as it correlates with different CEFR levels, as well as with the criteria observed by the expert teachers at the moment of labelling materials due to their complexity. Results also allowed to identify that vocabulary and syntax are the sub-aspects of complexity that are more predominant for students at different levels, and vocabulary and reference are the most observed aspects by teachers. The final objective of this research is to provide EFL teachers with a comprehensive framework which facilitates the measurement of texts’ complexity, which will help teachers provide their students with readings which meet their language needs.