Studying animal cognition is meaningful because it helps us understand how animals adapt to the natural environment. Many birds build nests, clean their nests and reject foreign objects from their nests, which provide an optimal opportunity for studying their cognition toward foreign objects in nests. However, hand-made models used in previous studies have many deficiencies that considerably constrain our capacity to understand the evolution of avian cognition of foreign objects because they are unquantifiable and dependent on different features. We established a 3D modeling and printing method to manipulate one geometric dimension of a model while controlling for others, which allowed us to investigate avian cognition for different dimensions independently. Here we introduce this method, conduct an empirical study as an example, and discuss its applications to further studies.