A Study On The Comprehensive Application Of Readability Formulae In EFL Context

Quite a few researches have been carried out recently to examine the validity and reliability of the classic readability formulae and in assessing the reading difficulty of the EFL text. However, those researches before focused on the individual application of the classic readability formulae and the comparison between the reading difficulty level results from the cloze tests and the readability formulae. A comprehensive application of different readability formulae are quiet needed in assessing the validity and reliability of the readability formulae in EFL field. Based on the previous study, the present research carries out a comprehensive analysis of six classic readability formulae in assessing the reading difficulty of20EFL texts. It aims to answer the following three research questions:(1) Does the comprehensive calculation of the reading ease score of the texts from the textbook give us reliable text difficulty level results that conform to the original text difficulty levels in the textbook?(2) Does the comprehensive application of readability formulae have different performances in four different language levels?(3) Does each of the readability formulae have different performances in assessing the reading difficulty level of the EFL texts? Among the six readability formulae, which one performs best and which one performs worst?20EFL reading texts that are divided into four language levels are involved in this research. The reading difficulty levels are decided by a comprehensive calculation through the six readability formulae and the new language levels of those texts are administered by ranking the reading ease score through the Main Divergence and Index Means. The results indicate that that there are some discrepancies between the original language level and the new language level calculated by the readability formulae through the calculating of the Index Means and Mean Divergence:It is also found that it is hard to reach an agreement for all the readability formulae. However, although there are many errors, we could see that more than half of the results from the readability formulae are correct. And we could also find that the majority of the incorrect predictions are only one level upper or lower than the final language level. Among the44incorrect predictions, only2of them are two levels away from the final language level. The assessment of the reading difficulty of EFL texts by the readability formulae is far from satisfaction. A comprehensive application of the readability formulae could only give us a rather rough guide of classification.