Examinations of Factors Which Produce a Spatial Clustering of Opinions in Dynamic Social Impact Theory

• Published in: Jikken shakai shinrigaku kenkyū Vol. 41; no. 1; pp. 16 - 25
• Main Authors: ISHIMORI, MASANORI, KOSUGI, KOJI, FUJISAWA, TAKASHI, MIZUTANI, SATOHIDE
• Format: Journal Article
• Language: English
• Published: The Japanese Group Dynamics Association 2001
• Subjects: a function transforming into binary; computer simulation; locally interaction; spatial clustering
• ISSN: 0387-7973; 1348-6276; 1348-6276
• DOI: 10.2130/jjesp.41.16

The purpose of this study is specifying the factors which produce a spatial clustering of opinion that had been consistently found in Dynamic Social Impact Theory (DSIT). In this study, we constructed four models for the purpose of examining the efficacy of each variable used in DSIT. In Faction-size model and Accumulative-influence model, each cell had two continual variables which indicated the strength of individual. Here, Model A and B had only one variable. Model C had no differences among individuals of strength and model D had a distance variable called“neighbor”. The results showed that each model produced a spatial clustering and that strength and the number of individual were not requirements. This result suggested that an internal function of transforming the output into binary variable and a locally interaction are necessary conditions to produce a spatial clustering. In this study, we suggested the direction of developing of DSIT, and pointed out the problems about a broad interpretation and misdirection of DSIT. Finally, it was argued that the problems which can occur in the process of constructing models in general by using computer simulation, and the perspectives of simulation studies in the future.