JOO, KI-IN; PHD

                         UNIVERSITY OF MINNESOTA, 1992

                         SOCIOLOGY, DEMOGRAPHY (0938); SOCIOLOGY, SOCIAL STRUCTURE AND DEVELOPMENT
 

                         In the context of sociology of science, scientists themselves have been often ignored by mainstream
                         concerns for understanding scientific activities and their institutions. Here the action agents of scientific
                         activities, scientists as categorically differentiable populations (including specialists in a broader sense),
                         become the focus of my analyses through their interaction structure. For this particular purpose, first
                         some plausible models are constructed (also modified and extended) from analogous ecological ideas
                         and related structural equation model. One of the main advantages in the model construction approach,
                         specially dealing with mathematical models, seems to appear when further analyses implies theoretical
                         extensions based on the speculated models. In addition, empirical predictions also becomes specified
                         from them. Every later development throughout the paper derived from or related to such model
                         construction approach. Theoretically, scientists' interaction pattern (competitive vs. cooperative through
                         different levels of interaction, horizontal or hierarchical) could be better understood by looking at their
                         interaction structure, assuming that the consistent structure, as a social system, emerges from interacting
                         populations. Accordingly, the issue of whether such emergent system is stable or not is raised.
                         Methodologically, the relevancy issue of traditional hypothesis testing on the interaction parameters is
                         first raised. Stimulated by such problem, the error distributions of the interaction parameters are
                         examined. While the effort to include more reality into my models was continued after the model
                         construction effort, most distinctive ones seem to comprise different resource inclusion effects (and also
                         different pattern of resource inclusion effect) into the model systems. Such model extension is
                         continuously considered throughout later simulation studies, which makes sensible when the population
                         dynamics is determined by the nonlinear differential structural equation model systems. Finally, to
                         support the utility of the above model approach, an empirical data set, 'Human Factors Specialists
                         Education and Utilization (1991),' is cautiously examined.

Social Systems Simulation Group P.O. Box 6904 San Diego, CA  92166-0904 Roland Werner, Principal Phone/FAX  (619) 660-1603