This study has three purposes: (1) to determine whether the graphic organizer (a schematic representation of text structure using key vocabulary terms) could be used to compensate for the effects of a passage organized with a description top-level structure; (2) to explore the effectiveness of providing students differing in reading ability with different types of instructional strategies and text structures; (3) to learn how graphic organizers, text structure, and reading comprehension level affect posttreatment attitudes. Tenth graders read two versions of an expository passage which contained identical information but differed in top-level structures (description versus comparison/contrast). Students were either exposed to or denied instruction with a graphic organizer that had been constructed to reflect the top-level structure of the comparison/contrast passage. Multiple regression analyses revealed these findings: first, immediate free recall performance was positively affected only under the description text condition; second, both skilled and unskilled readers appeared to benefit from the use of graphic organizers; third, text organized with a description top-level structure did not produce significantly better recall performance among poorer readers; and finally, multiple regression analyses of posttreatment attitude data did not yield any educationally significant results. (Author/GK)