属性层级模型的简介与述评

Introduction and Review of Attribute Hierarchy Method

属性层级模型是一种基于统计模式识别和分类的认知诊断模型。根据已确定的属性之间的层级关系,属性层级模型可以将被试在试题上的反应模式即观察反应模式划归为期望反应模式,以此得到被试的知识状态,进而能够对被试进行有效的、有针对性的补救。作为规则空间模型的重要变种,属性层级模型同样克服了标准测验理论单维性和只重视作答结果而忽视作答过程的考察分析的不足,且逻辑性更强,促进了教育和心理测量学的发展和分析。

Attribute Hierarchy Method （AHM） is a Cognitive Diagnostic Model （CDM） based on statistical pattern recognition and classification. According to the identified hierarchical ordering of attributes, Attribute Hierarchy Method will be assigned the observed examinee response patterns on items to the expected examinee response patterns, to be estimate examinees＇ attribute mastery patterns associated with different cognitive skills and further help design effective and targeted remedy for individual examinees. As an important variation of Tatsuoka＇s Rule Space Method （RSM）, AHM overcomes the Standard Test Theory＇s shortcomings as same of one- dimensional and only care the scores but ignore to inspect and analysis the process to finding answers; also, AHM is better in logic and promotes the development and analyses of educational and psychological tests.