An experiment was conducted to test the effects of isolating and identifying generalities and instances in written prose instruction on student performance. Eighty college students were given either a typical chapter from a college textbook that included a summary, the same chapter with the summary removed, a modified version of the chapter in which generalities and instances of those generalities were isolated and irrelevant material was eliminated, or no instruction. Isolating generalities improved performance on a 20-item memory and application level test. Isolating generalities and instances bettered efficiency, which was defined as test score divided by time spent. An unexpected interaction between treatment and the sex of subject occurred. Males exposed to the chapter with the summary had higher test scores than males exposed to the modified chapter. These results were reversed for females. The results provided support for the proposition that instruction designed to teach a given generality should be structured by isolating and identifying descriptive statements and instances of that generality. (Author/RL)