This study assessed the degree to which disabled readers commonly and identically confuse high-frequency sight words in isolation and context. Disabled readers from a university reading clinic and from fourth-grade classrooms were tested for their ability to recognize 22 target words in word lists and in a story. Series of analyses of variance indicated that both groups recognized more words in context than in isolation, and that the clinic and natural school groups did not differ in the number of errors produced in isolation or in context. A Spearman rank-correlation coefficient yielded a very low, statistically significant correlation between the number of errors in isolation and context for the 22 target words. Hierarchical arrangements of words confused in both conditions and words frequently confused in isolation, but rarely in context, are presented. Inspection of these errors suggests that readers relied on visual characteristics of words for isolated word recognition, whereas errors on these words in context were generally regulated by semantic and syntactic constraints of the story. Findings are discussed in terms of diagnosis and instruction through the use of context materials. Suggestions are given for futher research with learning disabled students in natural school settings to identify specific reversal problems with high-frequency words.
