CONTEXT EFFECTS ON WORD RECOGNITION AND READING-COMPREHENSION OF POORAND GOOD READERS - A TEST OF THE INTERACTIVE-COMPENSATORY HYPOTHESIS

THIS STUDY examined context effects on word recognition and reading co mprehension, testing Stanovich's (1980, 1984) interactive-compensatory model. Forty-eight good and poor third-grade readers orally mad easy and difficult passages that contained altered words created by graphem e substitution (i.e., changing a single letter to produce a word that differed in meaning from the original word) to permit examination of r elative reliance on semantic, contexual information and orthographic i nformation. Semantic context was manipulated by normal and randomly or dered presentation of the sentences. After oral reading, each student was given a recognition test of the target words (altered and original ) and a comprehension test. Overall, oral reading revealed the dominan ce of orthographic information in word recognition of both good and po or readers, and both semantic context and text difficulty influenced t he reading of target words. Good readers used orthographic information (i.e., read the altered words) more than did poor readers, whereas po or readers used semantic context more than good readers, a pattern of results generally supportive of the interactive-compensatory model.