摘要
传统英语阅读测评只呈现单一的终结性分数,分数背后是否存在学生对阅读理解强弱程度的掌握信息却不得而知。认知诊断评估可为学生提供多维度、细颗粒度的测评结果,以弥补传统单维分数成绩的不足。本研究运用认知诊断测评技术,结合树回归分析,采用量化与质化相结合的方法,旨在研究英语阅读认知诊断测评模型的构建。研究结果确定了8个英语阅读认知属性,并成功生成群体和个体层面的诊断信息,最终形成图文并茂的诊断反馈。研究结果有助于教师发现学生阅读的优势与不足,为个性化"量身定制"的教学提供强有力的指导。
Traditional EFL reading assessment merely provides students with a summative score, lacking diagnostic information about their strengths and weaknesses. The emergence of Cognitive Diagnostic Assessment(CDA) offers a better feedback with multi-dimensional and fine-grained results, thus overcoming the deficiency of classical tests. With this aim, this study constructed a cognitive diagnostic model for EFL reading test by combining CDA approach with Tree-based regression analysis and other statistical approaches. The results identified 8 reading attributes and successfully generated both grouplevel and individual-level diagnostic information. Presented in a well-designed diagnostic feedbackformat, students will easily find out their reading problems to promote their learning.
出处
《外语教学与研究》
CSSCI
北大核心
2018年第1期74-88,共15页
Foreign Language Teaching and Research
基金
国家社科基金项目"英语认知诊断测评模式构建及有效性论证"(17BYY015)的部分成果
孟亚茹主持的教育部人文社科基金项目"移动语言学习MALL与西部英语生态教学的整合模式构建-基于2006年来的历时研究"(12YJA740057)的资助
