Abstract
Incoming freshmen are typically required to write essays which are then holistically rated to determine composition course placement. These placement essays vary not only in topic, but also in the way the topic is structured. Two topic structures are most commonly used: Open (students draw on their own knowledge) and Response (students read a given text and respond to it). It has been established that students perform differently on topic structure itself. To investigate this effect, one topic was used but presented as (1) an Open topic structure, (2) a Response topic structure with one reading passage, and (3) a Response topic structure with three reading passages. The essays, written by college freshmen, were holistically rated for quality and analyzed for fluency, total error, and error ratios. The results indicated that the structure of the topic made a difference in quality, fluency, and total error, but not in any error ratio. These results suggest that, for placement testing, one should first decide which types of students one wishes to identify because each topic structure distinguishes low, average, and high ability students differently.