NORMALIZED EIGENVECTORS AND THE ANALYSIS OF PSYCHOLOGICAL DATA

DOI Web Site

Bibliographic Information

Other Title
  • 規準化固有ベクトルによる素得点データの分析
  • キジュンカ コユウ ベクトル ニ ヨル ソ トクテン データ ノ ブンセキ

Search this article

Description

In analyzing psychological data, it seems customary to use correlational or covariance matrices. However, we herein discuss the analyses of raw score type data matrices and of the inner product moment matrices based on them.<br>First of all, we define raw score type data which we are going to use, and discuss the algebraic decomposition of raw score type data matrices and of inner product moment ones. We show that each element of the first component for both matrices is related to the mean value of each item under a certain condition, and that it may play an important role in describing psychological phenomena.<br>Secondly, when the dimensions for data are over two, we show that the orthogonal transformation of normalized eigenvectors gives good interpretation, and that the rotated solution may be equivalent to a kind of factor scores.<br>Two numerical examples are presented. The one is concerned with the application of our method to the analysis of latent structure of the work-curve for the Uchida-Kraepelin test. The model for the work-curve has been proposed by Kashiwagi (1962). The other one is concerned with the evaluation of the scores of the Fatigue Scale which was proposed by the Fatigue Research Committee, Japan. And we found that the scale is very useful in the sense of its interpretability of fatigue phenomena.

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top