Children's recognition memory for written sentences: a comparison of good poor readers

In 18 pairs of fifth-grade males matched on sex, CA, and IQ, 1 was a good reader and 1 a poor reader. Each child read 8 3-sentence acquisition stories, subsequently performed a card-sorting task, and finally a recognition test consisting of 40 sentences which either were in the acquisition set or were 1 of 5 kinds of transformation. Retention was better for semantic aspects of the stories, but retention of tense was poor. Good readers mad fewer recognition errors than poor readers on new sentences which were true inferences from the acquisition list and on tests for retention of tense and number.