Quantitative Research on Readability Formula of Texts for Chinese Heritage Language

Abstract

The compilation of Chinese heritage textbooks has always adopted the standards of teaching Chinese as mother language or as second language. The difficulty level of the texts is mostly measured separately from the aspects of Chinese characters, vocabulary, and grammar. There is no comprehensive quantitative evaluation standard. Readability formula is one of the methods to comprehensively measure the difficulty of text. The present study made use of Chinese heritage textbooks as the source of data and constructed readability formula by the method of multiple regression analysis. The formula includes four language features, i.e., proportion of difficult words, number of different characters, average sentence length, and proportion of function words, which could explain 72.9% of the variation of text difficulty level. Limitations and future works of the formula are discussed.