Text‐related variables in textbook readability

Abstract

Abstract Text‐related variables such as sentence complexity and vocabulary load were examined in three secondary‐school human and social biology textbooks. Content and non‐content words, technical (including biological words) and non‐technical words, rare and frequent words, and word repetitions were considered. These were intercorrelated with readability estimates from teacher and student judgements, a cloze test, and a Fry graph to establish the significance of these variables in readability. The study involved 397 ‘secondary three’ students in ‘express’ classes from five, English‐stream, government secondary schools, and 85 biology teachers from 133 secondary schools in Singapore. Ten cloze tests with every fifth‐word deletions were administered to the students, and marked by exact‐word scoring. The response to six Likert‐type items on the language of the passages was elicited from students at the time of the cloze tests, and from teachers through a postal questionnaire. Word frequency in the cloze passages was established by the Standard Frequency Index. Words were classified into technical, non‐technical, and biological by reference to scientific dictionaries. Counts were made of the repetition of rare and frequent words, content (technical and non‐technical) words, and non‐content words. Sentence complexity was examined by mean sentence length and the Structural Index. The readability of these textbooks was found to be at the frustration reading level of the students. Sentence length was inadequate as a measure of syntactical complexity or textual difficulty. Of the variables examined, only teachers’ judgement on language correlated significantly with the cloze scores. The percentage of technical or biological words, and the repetition of content (technical and non‐technical) words were low. Student judgements on language were much influenced by technical or non‐technical words, and the repetition of technical content words and rare words. However, these variables could not account for the students’ cloze performance. Students judged the language of the passages as difficult when the repetition of technical content words and rare words was high.