Effects of Embedded Anomalies and Oral Reading Errors on Childen's Understanding of Stories

The basic question asked by this study was, “Do mistakes made while reading interfere with comprehension?” A secondary aim was to clarify the competing views of Gough (1972) who argues that the unskilled reader would not make hypotheses about what words in the text should be, and Goodman (1976a), who argues that the generation of tentative hypotheses about meaning is an inevitable part of the reading process. Three experiments were conducted. The first two were simulations in which skilled readers, about 9 years of age, read stories which varied according to embedded error type, difficulty, error rate, text access, and set strength. The third experiment was a naturalistic comparison study, using unskilled readers. Overall, the results offered support for both Gough and Goodman, depending on the kind of comprehension tested and the criterion for acceptability of responses. It seems that accuracy is necessary for atomistic precision; less so for global interpretation.