ABSTRACT
Many contemporary tests include both multiple-choice items and ex tended constructed-response items, for which the item scores are ordered categorical ratings provided by judges. Examples include the National As sessment of Educational Progress (NAEP) (Calderone, King, & Horkay, 1 997), many of the Advanced Placement (AP) examinations (College En trance Examination Board, 1 988), state assessments that have been ad ministered at times in North Carolina, Wisconsin, and in many other states, and other less visible testing programs. If the collection of items is sufficiently well represented by a unidimensional IRT model, scale scores may be a viable plan for scoring such a test.
