Evaluation of a hierarchical mapping technique as an aid to prose processing

Abstract

Seventeen college students were trained on a hierarchical mapping technique designed to facilitate prose processing. This technique required the students to transform text into node (concept)-link (relationship) networks using a set of experimenter-supplied links and structures. Following training the students used the mapping strategy in studying a 3000-word passage extracted from a geology textbook. Five days later they were given four types of tests over this article: multiple-choice, short-answer, essay, and a summary-oriented concept cloze. The treatment group's performance on these tests was compared with a control group ( N = 21) who used their normal study and test-taking techniques. A Hotelling T 2 test indicated that the treatment group significantly outperformed the control group on the dependent measures. Further analyses showed that the major differences between groups were attributable to the concept cloze and essay exams, both of which were designed to assess performance on “main ideas.” This pattern of results suggests that the strategy is valuable in assisting the student in the extraction and retention of main ideas but does not appear to affect the extraction and retention of details. Finally, 2 × 2 factorial analyses of high and low grade point average (GPA) subgroups indicated that the strategy may be more beneficial for low GPA students. Implications of the results are discussed in terms of organization theory and the encoding variability hypothesis.