Rating Scales and Rubrics in Language Assessment

Abstract

In language testing and assessment settings, language learners are often required to perform tasks which generate extended samples of speech and writing, such as interacting in discussions or writing essays. The more complex these samples are, the more they represent the complexity of language ability and use—and the more challenging they are to score. Whether they are for high‐stakes standardized examinations, or classroom‐based language performance assessments, rating scales are tools that are developed in an attempt to score these performances efficiently, systematically, and transparently. For decades, the term “rating scale” has been the preferred term. However, “rubric” is now used as often, if not more often, in research related to writing assessment in particular, and is a more familiar term to researchers and practitioners in educational settings. The provision of a rating scale and rater training does not automatically lead to similar scores among different raters with that scale (i.e., inter‐rater reliability). In addition, there currently exists a certain tension in research on rater–rating scale relationships: On the one hand, raters can be perceived as sources of potential measurement error that threaten quality rating, and, on the other hand, they can be perceived as resources and sources of expertise. This entry discusses rating scales as they are used in the assessment of language performance, and includes a discussion of their purposes and development, research on current issues, and challenges in their use.