Abstract
The spatial congruency bias refers to the phenomenon where people are more likely to judge two sequentially presented objects as being the same identity if they appear in the same location compared to in different locations (Golomb et al, 2014). Such a special role of spatial location in object identification has been demonstrated with diverse types of stimuli in previous research (e.g., abstract objects, colored squares, faces). However, the reasons why observers are prone to spatial location’s influence on object-feature judgements remain unknown. We propose two theoretical types of mechanisms underlying the spatial congruency bias: vision-based and knowledge-based. According to the vision-based hypothesis, location is fundamental (and uniquely prioritized) in visual processing and its privileged role should be evident early in development. However, if the spatial congruency bias is knowledge-based, it should emerge during development as children acquire knowledge that objects in the same location are unlikely to change identity. To distinguish between these possibilities, we conducted a developmental study with young children. In three experiments of varying difficulty, adults or 5-year-old children were presented with two sequentially presented abstract objects and were asked to make same/different identity judgments. We replicated the robust spatial congruency bias in adults, but critically, 5-year-old children did not exhibit the bias despite exhibiting similar discrimination accuracy to the adults. The results suggest that spatial congruency bias is a product of development, perhaps resulting from learning that in the real world the same location generally cannot be occupied by different objects and that stable objects are unlikely to suddenly change their identities. Thus, the spatial congruency bias may reflect a robust example of how implicitly learned assumptions about the world can influence visual object recognition.