The purpose of the present study was to explore the effect of graphic organizer instruction versus outlining on students' text recall in tenth-grade world history. Based on the hierarchical structure of graphic organizers depicting interrelationships among ideas, we hypothesized that students instructed in the use of graphic organizers would display significantly higher text recall than a group using outlines. Seventy-two tenth graders in three sections of world history participated in the study. Two sections received systematic instruction in the development of graphic organizers and a third section received parallel instruction in outlining. One of the two graphic organizer sections had previous instruction in summarization and question generation during an 154Journal of Reading Behavior earlier study. Six 15-item multiple-choice quizzes comprised one of the dependent measures. On the first five quizzes there were no significant differences between the three sections. However, when teacher prompting was phased-out on quiz six, the graphic organizer group with previous training in summarization significantly outperformed the other two groups. Additionally, on a second dependent measure involving the development of a written recall protocol based on difficult college level text, the group with previous summarization experience significantly outperformed both the new graphic organizer group and the outlining section. Implications for a cumulative, long-range view of metacognitive instruction are discussed.

The purpose of this study was to test the hypothesis that training in the construction and use of hierarchically organized knowledge would enable students to make predictions across chapters in a 10th-grade world history text. First, a history text that provided only exemplars of the American, French, and Russian revolutions was modified by the insertion of explanatory material on the causes of revolutions. Then, two 10th-grade classes were randomly assigned to experimental and control group status. The experimental group had instruction in: (a) the construction of hierarchically organized knowledge via the use of graphic organizers, and (b) use of this knowledge for predicting events. The control group received traditional instruction. They read the modified text, outlined it, and engaged in regular class discussions of the material. The results revealed that on regular chapter tests both groups displayed equal comprehension of textbased information. However, on an essay test asking students to predict events in the Cuban Revolution which they had not read about in their history text, the experimental group significantly outperformed the control group. The results indicate that textbook explication alone of causes of events is insufficient for acquisition and use of hierarchically organized knowledge. Students must be taught how to construct text-based information into hierarchically structured form and use it for making predictions.A reader's prior knowledge or schemata can provide the foundation for making predictions about oncoming information in a text (Anderson, 1984;Anderson & Pearson, 1984). Schema theory postulates that our knowledge is represented in memory in a hierarchical structure such that macrolevel information guides our interpretation of lower-level events and details (Rumelhart, 1984;van Dijk & Kintsch, 1983).One of the limitations of schema theory is that it does not specify how schemata are constructed (Bransford, 1984). Schema theory simply draws upon general gestalt theory to suggest that readers spontaneously organize their knowledge. Spontaneous organization requires a reader to abstract from a series of exemplars their common properties and group them into a superordinate schema. Presumably, this superordinate schema would then automatically be instantiated by a reader to subsume similar information and to infer missing slots in a text through default assignment.However, simply possessing prior knowledge is no assurance it will be activated. Other reader-based factors influence the degree to which a reader will use existing knowledge to predict oncoming ideas in a text. For example, a reader who learns to organize knowledge hierarchically rather than in a linear list is able to perform an efficient search of memory to generate predictions (Voss, 1984). In addition, procedural knowledge is required in order to use prior knowledge efficiently (Just & Carpenter, 1984). The mental procedures involved in organizing information hierarchically and in using this information to make predictions m...