Een datareductiemodel voor relationele gegevens met behulp van cumulatieve matrices

Samenvatting

Factor analysis and cluster analysis, as well as various scale models are well-known devices for extracting tendencies and characteristics from complex attitudinal data. An attempt is made to use cumulative matrices to develop a data reduction model for relational data. A sociomatrix is a binary matrix which is rectangular, with the value n representing the number of r's. The mechanism for extracting single or multiple solutions from this type of matrix is given. It is applied to an investigation into the formation of subgroups in a Franciscan community in De Bilt, the Netherlands. The method proved useful in sociometric analyses, power and influence relations, and other phenomena which involve multiple criteria. P. Tiersma