《中国英语能力等级量表》阅读策略描述语效度验证——Rasch测量模型视角

Validating descriptors of the reading strategic competence subscale of CSE:Rasch measurement model perspective

本研究从Rasch测量模型视角,通过全国范围的问卷调查及其结果分析,验证了《中国英语能力等级量表》阅读策略描述语在高等教育阶段的效度。研究结果表明,阅读策略描述语拟合度较好,总体上描述语的难度排序和级别划分合理。虽然少量描述语呈现DIF,描述语的难度排序和级别划分数量与专家预测略有不同,但研究在效度验证的基础上探寻了原因并对描述语作出相应删改,确保了描述语质量,使描述语的级别代表性得到优化。

Through a nation-wide questionnaire survey and Rasch analysis,this study validates the descriptors pertinent to the higher education stage in the reading strategic competence subscale of China’s Standards of English Language Ability(CSE).Results show a moderate goodness-of-fit of the descriptors,indicating that they are reasonably well-scaled in terms of difficulty.Though a few descriptors are flagged as DIF items,and the hierarchical order and the number of descriptor levels are slightly different from the prediction of the experts,they are either deleted or revised,based on the evidence from the validation.The quality of the descriptors is thereby ensured,with their level-distinctiveness enhanced.