Complexity and decision making

This article applies to decision making at the level of the individual human organism. Observations of real-world systems and simulations of them have suggested that many decision-making environments are extremely complex. Indeed, given the limitations on human information processing capacity derived from laboratory experiments, we should not be able to operate in them at all. The methodology developed allowed the controlled investigation of the interaction of the individuals with dynamic systems. The effect of three complexifying factors on a subject's ability to optimize total welfare in a series of computer models of a welfare administration project over time was studied. The factors of (i) number of elements in the system, (ii) connections between them and (iii) the presence or absence of random variation were realized in a 2 × 2 × 2 analysis of variance research design in which two male and two female undergraduates served as subjects in each cell. Multivariate analyses of variance performed on derived measures of performance and system control demonstrated either insignificant differences, or significant differences in the direction opposite to that predicted. The more complex systems did not always attain poorer levels of performance than simpler systems. The results suggest that revision of theories of complexity within the social sciences may be necessary.