字母混淆矩阵的多维尺度分析

Multidimensional Scaling Analysis of Letters Confusion Matrix

为了阐明多维尺度分析法(MDS)在认知心理学领域的独特适用性和恰当性,简要介绍MDS的原理和数据类型,再运用MDS程序对Mayzner(1965)有关字母混淆矩阵的认知心理学早期研究案例进行再分析。结果表明,MDS对认知心理学研究间接数据的处理是恰当的,可以用来进一步挖掘混淆矩阵的数据信息。而且字母混淆矩阵的再分析结果发现,快速呈现字母条件下,影响被试辨认字母的因素不仅是字母的视觉特征,可能还有字母的书写动作特征。

To illuminate the unique applicability and appropriateness of multidimensional scaling method（MDS） in cognitive psychology domains, a brief introduction regarding the principle and data type of MDS are provided. Then a early case of cognitive psychology re- garding letters confusion matrix studied by M. S. Mayzner（ 1965 ）is reanalyzed via MDS procedure. The results verify that MDS is appro- pilate for secondary data of cognitive psychology and can be used to mine confusion matrix data. And the reanalysis of letters confusion matrix indicates that not only visual features but also writing movement characteristic of letter influence participants to identify the letter presented quickly.