Abstract
These studies investigated the degree to which prompts and topic types affect the writing performance of college freshmen. The students (N = 3,452) taking the 1989 and 1990 Manoa Writing Placement Examination (MWPE) were required to write in response to two types of topics (for a total of 6,904 essays): one in response to a reading passage and another in response to a question based on personal experience. Ten such prompt sets were used in this study. Study 1 indicated that the MWPE testing procedures were reasonably reliable and consistent across semesters but that student responses to individual prompts and prompt sets were significantly different from each other. Study 2 showed that if two topic types and a large number of prompts are involved, the differences that arise in the performance on prompts or topic types can be minimized by examining the students' mean scores and changing the pairings so that the prompt sets are more equitable in subsequent administrations.