Automated Complexity Assessment of English Informational Texts for EFL Pre-service Teachers and Translators

Abstract

Automated text complexity assessment tools are of enormous practical value in solving the time-consuming task of analyzing English informational texts for their complexity at the pre-reading stage. The present study depicts the application of the automated text analysis system the TextEvaluator as an effective tool that helps analyze texts on eight dimensions of text complexity as follows: syntactic complexity; academic vocabulary; word unfamiliarity; word concreteness; lexical cohesion; interactive style; level of argumentation; degree of narrativity, with further summarizing them with an overall genre-dependent complexity score. This research examines the complexity dimensions of English informational texts of four genres – legal, linguistic, news, and medical – that are used for teaching reading comprehension to EFL (English as a foreign language) pre-service teachers and translators at universities in Ukraine. The data obtained with the help of the TextEvaluator has shown that English legal texts are the most difficult for reading comprehension in comparison to linguistic, news, and medical texts. In contrast, medical texts are the least challenging out of the four genres compared. The TextEvaluator has provided insight into the complexity of English informational texts across their different genres that would be useful for assembling the corpora of reading passages scaled on specific dimensions of text complexity that predict text difficulty to EFL pre-service teachers and translators.

Keywords: automated сomplexity assessment, informational texts, text complexity, text complexity indices, the TextEvaluator, EFL pre-service teachers and translators.