2021-12-06T11:36:38Zhttps://nagoya.repo.nii.ac.jp/oai
oai:nagoya.repo.nii.ac.jp:000022662021-03-01T21:23:55Z
EXPLORATORY FACTOR ANALYSIS OF MULTISET DATA多集合データのための探索的因子分析村上, 隆6077MURAKAMI, Takashi6078A new exploratory method, multiset factor analysis, which factor analyzes data consisting of several sets of variables related to different domains or levels was proposed. Basic equations of this method are written as [numerical formula] where R_<kk'> is n_k by n_k correlation matrix between variables of set k and k', A_k is n_k by q_k factor loading matrix for set k, S_<kk'> is q_k by q_<k'> correlation matrix between factor scores satisfying [numerical formula] and D_<kk'> is diagonal matrix including unique variances when k = k', and zero matrix when k ≠ k' An alternating least squares algorithm is formulated. It assumes that [numerical formula] and [numerical formula] where L_k is diagonal and positive definite. Then following three equations [numerical formula] [numerical formula] and [numerical formula] are alternated until convergence is reached. Final solutions satisfying (2) are obtained through the transformations [numerical formula] and [numerical formula] This algorithm is a natural extension of ordinary principal factor method. A questionnaire data collected from housewives in Aichi prefecture was analyzed by this method. lt consists of four sets, - buying behavior, attitude for buying, need for information and confidence for the various media. A factor which could not be found by separate factor analysis was derived in the first set and interesting relations between the factor and factors of other sets were revealed.国立情報学研究所で電子化したコンテンツを使用している。departmental bulletin paper名古屋大学教育学部1985-12-20application/pdf名古屋大學教育學部紀要. 教育心理学科327593http://hdl.handle.net/2237/368903874796https://nagoya.repo.nii.ac.jp/record/2266/files/KJ00000725981.pdfjpn