Authors
- Jason Fan, Deputy Director and Senior Research Fellow, Language Testing Research Centre University of Melbourne, Australia
- Tan Jin, Professor, School of Foreign Studies, South China Normal University, China
- Ivy Chen, Research Fellow, Language Testing Research Centre, University of Melbourne, Australia
Abstract
The Lexile Framework is an innovative approach to reading comprehension whose validity has been proven through over 20 years of ongoing research. The use of the Lexile Framework has greatly facilitated the teaching, learning and assessment of reading, which is considered as a vitally essential language skill for success at all levels of education. Despite the extensive recognition of its validity and use in numerous contexts, the Lexile Framework is not widely known in China; nor are the Lexile measures familiar to English language educators, learners and assessment developers in China.
This study aimed to explore the extent to which the Lexile Framework was applicable to the Chinese educational context by analysing the Lexile measures of the reading texts in four prominent, high-stakes English tests in China, representing reading demands at four distinct educational stages: junior secondary, senior secondary, collegiate basic and collegiate intermediate. The four English tests are: the English component of the Senior High School Entrance Examination (SHSEE); the National Matriculation English Test (NMET); and the College English Test Band 4 (CET4) and Band 6 (CET6).
To further interrogate the accuracy of Lexile measures in capturing the text complexity of the test materials, we also compared the text complexity results generated by the Lexile Analyzer, which implements the Lexile Framework, and Eng-Editor, a locally developed text complexity evaluation tool.
Two research questions were investigated in this study: 1) What is the range of the Lexile measures for the reading texts in English test papers at each of the four educational levels in China? 2) In what ways are Lexile measures comparable to text complexity estimates generated by Eng-Editor? To address the two research questions, a total of 800 benchmark texts, with 200 at each level, were randomly selected from a corpus which consisted of reading texts from the four high-stakes English tests in focus. These benchmark texts were then analyzed with the Lexile Analyzer and Eng-Editor respectively, which generated text complexity measures across multiple dimensions for each text. Multiple statistical procedures were implemented, including descriptive statistics, ANOVA, correlation analysis, ordinal logistic regression analysis, and linear regression analysis.
The results indicate that the Lexile measures of the benchmark texts could effectively determine their classification into the four educational levels in focus; there was a strong relationship between the Lexile measures of the benchmark texts and the corresponding complexity estimates produced by Eng-Editor. The findings of this study provide empirical support to the applicability of the Lexile Framework to English language education in China. Meanwhile, through the application of the Lexile Framework, this study also sheds light on the English reading demands of China’s four educational levels in the global context.