ABSTRACT
This chapter discusses features of essays based on multiple documents that are important to assess as delineated by theories of text comprehension and task-oriented reading. It focuses on situations in which essays are based on a preselected set of documents rather than situations in which students find their own texts. The chapter also discusses promising approaches to computer-based assessment of essays, and what is needed to develop and test systems specifically designed for essays based on multiple documents. It identifies several challenges for developing these systems that are grounded in theory, research, and practical problems associated with automatic assessment of the use of multiple documents in essays. The chapter presents research on existing scoring systems that are illustrative of these approaches, but it is important to note that this area of research is in its early stages and there are only a few studies that involve the automatic scoring of essays.
