Spatial Reasoning

We carried out two experiments to investigate how people reason about the spatial relations among objects. The experiments were designed to test a theory of spatial inference based on mental models. The theory predicts that problems requiring only one model of the spatial layout to be constructed should be easier than those requiring more than one model to be constructed, even when the multiple-model problems have valid conclusions. A contrasting theory, based on rules of inference, predicts that problems based on fewer inferential steps should be easier than problems based on more steps. The first experiment held constant the number of inferential steps specified by the inference-rule theory, but varied the number of models required to make a valid response: the one-model problems were reliably easier than the problems requiring more than one model. The second experiment contrasted opposing predictions from the two theories, and once again the results supported the model-based theory.