Principal component procedure in factor analysis and robustness

African Journal of Mathematics and Computer Science Research

Joy Chioma Nwabueze¹ 

关键词: Principal component;    factor analysis;    robustness;    random variables;    distributions.;   

DOI  :  

学科分类：计算机科学（综合）

来源: Academic Journals

【 摘 要 】

Principal component procedure has been widely used in factor analysis as a data reduction procedure. The estimation of the covariance and correlation matrix in factor analysis using principal component procedure is strongly influenced by outliers. This study investigates the robustness of principal component procedure in factor analysis by generating random variables from five different distributions which are used to determine the common and specific factors in factors analysis using principal component procedure. The results revealed that the variance of the first factor was widely distributed from distribution to distribution ranging from 0.6730 to 5.9352. The contribution of the first factor to the total variance varied widely from 15 to 98%. We conclude that the principal component procedure is not robust in factor analysis.

【 授权许可】

CC BY