摘要
心理数据具有秩统计量、总体分布未知、数据模糊性等特征,利用秩检验、Fisher判别以及人工神经网络等非参数统计方法对心理行为进行"画像"与分析,不仅可以勾画出仿真程度较高的心理行为"画像",而且可以降低对心理行为分类、判别与预测时可能存在的"失真"危险。在基于6732份数据的认知采择水平的言语认知画像模拟实例中,证实了动词、方位词等言语认知投射认知采择心理行为的"画像"效应。
Psychological data has some characteristics,such as rank statistics,overall distribution unknown,data ambiguity,etc.Using the non-parametric statistical methods such as rank test,Fisher discriminant and artificial neural network to perform"portrait"and analysis of psychological behavior,not only can the"portrait"of psychological behavior with high degree of simulation be sketched,but also the risk of"distortion"that may exist when classifying,judging and predicting psychological behavior can be reduced.In the case of speech cognition portrait simulation based on the viewpoint cognitive selection level of 6732 data,the"portrait"effect of speech cognition of verb,position word mapping viewpoint cognitive selection level is confirmed.
作者
徐芃
熊泽蔚
熊健
XU Peng;XIONG Ze-wei;XIONG Jian(Vocational Counsel Centre,Guangzhou University,Guangzhou 510006,China;BGI-Shenzhen,Shenzhen 518083,China;School of Economics&Statistics,Guangzhou University,Guangzhou 510006,China)
出处
《数理统计与管理》
CSSCI
北大核心
2021年第1期78-92,共15页
Journal of Applied Statistics and Management
基金
国家社科一般项目(18BYY082)
教育部人文社科一般项目(17YJAZH098)。
关键词
非参数统计
用户画像
心理行为
言语认知
nonparametric statistics
user portrait
psychological behavior
speech cognition