THE EFFECTS OF READER CHARACTERISTICS, TEXT FEATURES, AND COMPREHENSION PROCESSES ON READING COMPREHENSION

When the psychological processes underlying test performance are understood, psychometric theory dictates that item difficulty can be explained through these processes. The goal of the project was to improve understanding of reading comprehension and the role of reader characteristics, passage features, and comprehension processes in understanding text through the application of explanatory item response models. Participants were 1,190 students from 11 to 20 years recruited from eight schools within four districts. Students represented a range of reading comprehension skills in terms of their word reading, semantic language and operational (i.e., inference making, working memory capacity) skills. Measures included the grade appropriate Gates-MacGinitie reading comprehension and vocabulary subtests, letter word identification and numbers reversed subtests of the Woodcock & Johnson tests of cognitive abilities, test of word reading efficiency, as well as a researcher-developed test of background knowledge. The results indicated that reader characteristics including vocabulary, background knowledge, working memory and reading fluency were the most influential in explaining variation in reading comprehension item performance. Passage features explained some variation in item difficulties, with expository passages and deep cohesion being the most influential. Most importantly, a few text-reader interactions affected reading comprehension test scores. However, their effects were not pronounced, as good readers tended to perform better than poor readers regardless of the text they read. Better word and world knowledge was found to be the most helpful in understanding texts of variable difficulty. These findings are consistent with research that targets the building of vocabulary skill and background knowledge in order to improve reading comprehension, and suggests that the benefits of such development would apply to a wide variety of texts and to both memory for what has been read as well as drawing inferences from the text. The study further showed that explanatory item response models can be applied in a meaningful way to operational standardized tests, while also highlighting the limitations inherent in such application for explicating the general effects of text characteristics and reader abilities on the comprehension of written language.