蕴涵量表法在HSK阅读理解测验公平性研究中的应用

Using Implicational Scaling Procedure to Detect the Differential Passage Difficulty Order on the Reading Comprehension Test of HSK

阅读理解能力测验中所选择的文章在内容方面对不同专业背景的考生亚团体是否具有公平性的问题,是测验效度高低的重要证据,也是测验效度验证(validation)的重要环节。本研究以中国语言与文学专业考生为目标组,分别将经济学专业和生物医学专业考生作为参照组,采用效标测量和蕴涵量表分析相结合的方法,对HSK(高等)阅读理解测验的文章难度对三个不同专业背景的考生组的公平性问题进行了检验。研究结果表明,两个参照组考生尽管具有各自的相对专业优势,但他们在六篇阅读材料上获得的难度排列顺序与目标组考生完全一致;虽然目标组考生不具备汉语知识以外的其他专业优势,但因为HSK考试所选择的阅读材料没有涉及语言知识本身以外的特殊专业要求,因而测验对三个不同专业背景的考生具有较高的公平性。

As a state-level standardized test, the Chinese Proficiency Test （HSK） is designed to measure the general language proficiency of those whose native language is not Chinese. But within these groups there may be subgroups that differ in ways other than the language ability of interest. These differences may affect their test performance, and hence the validity of inferences made on the basis of test scores. When this systematic difference in test performance occurs that appear to be associated with characteristics not logically related to the ability in question, we must fully investigate the possibility that the test is biased. This research, based on the evaluation of examinees＇ reading proficiency levels through their teachers,using implicational scaling procedure, specifically addresses itself to the investigation of fairness for three subgroups： Chinese and Chinese literature group, economics and business group, and biology and medicine group, on the reading comprehension test of HSK. The empirical research indicates that the difficulty order of six passages is completely same across the three subgroups, the implicational scaling procedure could be used to detect differential item or passage functioning.