COMPUTERIZED SCORING? A QUESTION OF THEORY AND PRACTICE

<p>This article examines the implications of computerized scoring of placement essays for the theory and practice of writing assessment, particularly for the complexly interrelated issues of economics, universality, and validity; and then considers its broader implications for the theory and practice of writing instruction. It argues that the very feature that would make computerized scoring inexpensive—its generalizability for widespread use—undermines its validity. The ultimate criterion for measuring the validity of any placement instrument is whether the instrument matches the specific local conditions, in this case, the purposes and content of the courses in which the students begin writing instruction. A generalized scoring system by definition cannot meet that criterion. More importantly, a computerized scoring system undermines the hard-won though still insecure recognition that writing is a communicative act inseparable from audiences, purposes, and contexts.</p>