A preliminary study of cloze procedure as a tool for estimating English readability for Russian students

Abstract

This preliminary study examines the relationships between each of six first language (L1) readability indexes and the cloze passage mean performances of Russian EFL students. The cloze passages were created by randomly selecting 50 text passages from an American public library and deleting every 12th word in each passage to create a 30-item cloze procedure. The participants were 5170 EFL students from 38 universities in the Russian Federation. Each student was randomly assigned to take one of the 30-item cloze passages. The L1 readability indexes calculated for each of the 50 passages were the Flesch, Flesch-Kincaid, Fry, Gunning, Fog, and modified Gunning-Fog indexes. The preliminary results indicate that the L1 readability indexes were moderately to highly correlated with each other, but only somewhat correlated with the mean performances of Russian university students on cloze versions of those same passages. These results are discussed in terms of why the L1 readability indexes are moderately to highly correlated with each other but only somewhat correlated to the Russian EFL means. The authors also explain what they are planning in terms of further linguistic analyses (e.g., of variables like average word length, percent of function words, number of syllables per sentence, number of words per paragraph, frequencies of words in the passages, and so forth) and statistical analyses (including at least factor analysis, multiple regression analyses, and structural equation modeling) of these data.