一种基于Q矩阵理论朴素的认知诊断方法

发布时间：2018-04-23 04:31

本文选题：GDD + Q矩阵　；参考：《心理学报》2015年02期

【摘要】：现有的认知诊断方法均是在复杂的统计测量学知识基础上构建的,需要经过大量的运算才可实现对被试的诊断分类。这使得相关研究者及一线教师在理解和运用某一认知诊断方法时困难重重。相比之下,孙佳楠、张淑梅、辛涛和包钰(2011)提出的广义距离判别法(GDD)较其他认知诊断方法更简单易用且分类准确率高。本研究在改进的Q矩阵理论(丁树良,祝玉芳,林海菁,蔡艳,2009;丁树良,杨淑群,汪文义,2010)的基础上,借鉴GDD的思路,提出一种无需进行参数估计的朴素的认知诊断方法,即海明距离判别法(HDD)。根据判别方式的不同将其分为R方法和B方法。采用Monte Carlo模拟的研究方法,以模式判准率(PMR)和属性平均判准率(AAMR)作为衡量被试知识状态分类准确率的指标,与GDD进行比较。结果表明,HDD具有更简便的操作步骤和更好的分类准确率。
[Abstract]:The existing cognitive diagnosis methods are all constructed on the basis of complicated statistical measurement knowledge and need a lot of calculation to realize the diagnosis classification of the subjects. This makes it difficult for researchers and teachers to understand and apply a cognitive diagnosis method. In contrast, the generalized distance discriminant method proposed by Sun Jiannan, Zhang Shumei, Xin Tao and Bao Yuqian (2011) is simpler and easier to use than other cognitive diagnosis methods, and the classification accuracy is higher. On the basis of the improved Q matrix theory (Ding Shuliang, Zhu Yufang, Lin Haijing, Cai Yancheng 2009; Ding Shuliang, Yang Shuqun, Wang Wenyi 2010), a simple cognitive diagnosis method without parameter estimation is proposed. That is to say, the method of distance discrimination of Hemin is HDD. It can be divided into R method and B method according to the different discriminant ways. The Monte Carlo simulation method was used to evaluate the classification accuracy of knowledge status by using the model recognition rate (PMR) and the attribute average rate (AAM) as the index to evaluate the classification accuracy of the knowledge status. The results were compared with GDD. The results show that HDD has more simple operation and better classification accuracy.
【作者单位】：江西师范大学心理学院;
【基金】：江西省研究生创新专项基金(YC2013-B024)资助
【分类号】：B842.1

【参考文献】