The Effect Of Rater’s Familiarity With Test-Takers On The Oral Scoring

With the development of China, more and more families tend to send their children to study in the overseas universities, since there exists fierce domestic competition for entrance examination of university, and also the wonderful and advanced overseas educational resources. Therefore, oral English have been the focus of schools, students and families. To evaluate test-takers’oral performance, oral test is the best way. However, different from other English skill test, such as listening test, oral test is much more subjective, especially in the rating process. Raters will evaluate the test-takers’performance according to their own interpretation of rating standards, which may be influenced by many factors, such as raters’age, educational and experience background, nationality, gender, physical conditions etc. What should be noticed here is that most of the researches focus on the interaction between raters and interaction between raters and rating-scale, few of them pay attention to the interaction of raters and test-takers, which may affect the oral test reliability. It is very common that raters of the oral test have different familiarity with test-takers, therefore, different familiarity may cause the scoring difference. In this study, A questionnaire-based and oral presentation experimental test was designed in this thesis to explore the relationship between the raters’familiarity with test-takers and the scoring of participants’oral English performance. The18participants were divided into two groups:Group Familiar and Group Unfamiliar, based on the questionnaire result of one rater who was their English teacher. After that, all of them participated in the topic-based oral test which would be evaluated by two raters, one to be their own teacher and the other one not. Independent Sample T Test was used to compare two groups’scores given by two raters, and it was found that scores of Group Familiar given by the rater (their teachers) was8%higher than those given by the other rater. The result of the thesis showed that raters with high familiarity with test-takers would score higher.