ABSTRACT
This special issue was motivated by the move from research to operations for computerized delivery and scoring of complex constructed response items. The four papers presented provide an overview of the state of the art for such applications. The issue begins by describing the range of computer delivered formats and computerized scoring systems that are currently in use. The remaining papers provide three views of validity in the context of computer delivered and scoring assessments. It is hoped that together, these articles will provide the reader with both an appreciation of the state of the art for computer-automated scoring systems, as well as a perspective on the issues that must be considered and the evidence that must be collected to produce automated scoring systems that allow for valid inference.
