Lexical diversity as a predictor of complexity in textbooks on the Russian language

The parametric model of the text as a research problem is of paramount importance in modern linguistics and education, since it opens up new approaches to understanding the processes of comprehending texts of various types. In the current study, 17 Russian language textbooks for elementary school were employed to identify correlations between lexical diversity indices and other complexity predictors. The total volume of the corpus compiled for the study is 439,938 words. The two-stage research algorithm included the evaluation of the reference values of text features at the basic level (word length, sentence length, the number of unique, non-repeating words and the number of word forms), evaluation and subsequent contrasting of complexity predictors, i.e. lexical diversity and readability indices. All calculations were performed with the automatic text analyzer RuLingva. The study revealed a positive dynamic of readability and no evidence of lexical diversity increase across grades. An average level of vocabulary diversity and overlaps of every 4th word in the text are fixed. No indication of correlation between text readability and lexical diversity is found. The obtained results can be useful to researchers, textbook authors, and teachers selecting textbooks. The prospects are seen in implementing functional and epidigmatic stratification of the vocabulary of the Russian textbooks under study.