Abstract
The purpose of this study was to explore opposing theoretical viewpoints as they relate to and attempt to explain the effects of text underlining. The subjects were 67 provisionally admitted freshmen who were randomly assigned to one of four experimenter‐generated underlining or underlining and annotating conditions, or a fifth group who generated their own text marking. During two sessions, all subjects took a test of prior knowledge, read the assigned passage, and took a 24 item multiple‐choice test consisting of 12 high relevant and 12 low relevant questions. Data analysis supported the von Restorff effect as a theoretical explanation of text underlining since subjects in the high relevant groups answered more high relevant items correctly, while subjects in the low relevant groups answered more low relevant items correctly. Subjects who generated their own underlining did not perform significantly better than those who were given experimenter‐generated marking. Additionally, prior knowledge was not significantly related to the three dependent measures.