This article discusses experimental work with computer-simulated microworlds as a means for overcoming the tension between laboratory research and field research in psychology. Research with such microworlds involves the study of how subjects interact with complex computer simulations of real systems, such as a small town or a forest fire. This kind of research has followed three different lines: (a) the individual differences approach, (b) the case study approach, and (c) the experimental approach. In the former approach, large groups of subjects interact with a given system, and the differences in their behavior is noted and related to various measures of abilities and personality. In the second, the behavior of individual subjects is analyzed in great detail to find typical and atypical behavior in an attempt to generate hypotheses. In the third approach, the effects of various system characteristics on behavior are studied at the group level. Experiments with such microworlds differ in important respects from traditional psychological experiments, and come closer to the situation studied in fieldwork.
