Research on Controlling Central Rating in Net-based Scoring of Subjective Test Items
Hengli Peng · Zhongguo kaoshi · 2013
Researchers find that in scoring process of subjective test items such as composition scoring, some raters tend to assign central scores, avoiding using the high or low end of the rating scale. This is called central rating . Central rating belongs to systematic scoring errors which will cause deterioration of test quality. This paper analyzes central rating in net-based scoring of subjective test items. It suggests three types of central rating and analyzes causes of central rating. Central rating can be detected by two methods, which are anchor paper method and statistic index method. All through the test development and scoring process, multiple measures should be applied to control central rating.