The Effects of Simulation Complexity and Hypothesis-Generation Strategy on Learning

This study investigated learning with computer simulation as a function of two factors: learners’ hypothesis-generation strategy and simulation presentation format. Simulation presentation format refers to whether the simulation was presented in its most complex form initially or whether it was presented in sections of increasing complexity. There were three main findings. First, there was an interaction between hypothesis-generation strategy (generating one versus several initial hypotheses) and complexity. Breaking the simulation into sections of increasing complexity did not increase overall task success but did improve performance on the section of lowest complexity for subjects generating several hypotheses. When complexity was low, subjects were more likely to generate several hypotheses. Second, generating several hypotheses was helpful when the simulation contained a variable that behaved counterintuitively. Third, when complexity was low, subjects in the multiple-hypothesis condition attempted to eliminate hypotheses more frequently than those in the single-hypothesis condition.