Two experiments examined whether different task ecologies influenced insight problem solving. The 17 animals problem was employed, a pure insight problem. Its initial formulation encourages the application of a direct arithmetic solution, but its solution requires the spatial arrangement of sets involving some degree of overlap. Participants were randomly allocated to either a tablet condition where they could use a stylus and an electronic tablet to sketch a solution or a model building condition where participants were given material with which to build enclosures and figurines. In both experiments, participants were much more likely to develop a working solution in the model building condition. The difference in performance elicited by different task ecologies was unrelated to individual differences in working memory, actively open-minded thinking, or need for cognition (Experiment 1), although individual differences in creativity were correlated with problem solving success in Experiment 2. The discussion focuses on the implications of these findings for the prevailing metatheoretical commitment to methodological individualism that places the individual as the ontological locus of cognition.