Emily Westrick¹, Benjamin Balas¹; ¹North Dakota State University

Face pareidolia research has contributed to our understanding of how face-like features are organized and identified throughout childhood and into adulthood. Most pareidolia research has either examined participants’ responses to images with features similar to the 1st-order configuration of faces, or asked participants to identify such features in random or textured patterns. McCloud (1992) identified another aspect of face pareidolia that specifically emphasizes top-down processes and perceptual organization: Placing a single eye within arbitrary silhouettes to impose face-like appearance on random shapes. His claim was that eye placement is sufficient to impose face-ness on such closed forms, with participants re-organizing and re-interpreting silhouettes as faces to accommodate the eye micro-pattern. Presently, we used this unique type of face pareidolia to ask whether children and adults exhibit systematic preferences regarding where an eye should be placed in an arbitrary shape and whether these preferences are differentially affected by broad spatial biases and shape-based biases as a function of age. We presented children (#N=137) and adults (#N=47) with four ambiguous silhouettes and prompted them to place one eye sticker on each shape wherever they felt was best. Across participant groups, these silhouettes were presented either in their original orientation, or flipped upside-down. Based on prior research describing young children’s tendency to disregard orientation relative to paper borders when drawing human figures (Goodnow and Friedman 1972), we hypothesized that younger children would place eyes in a more shape-dependent manner (being less affected by silhouette orientation), and older children would share a tendency with adults for more space-dependent placement. Our analyses revealed just such an interaction between age group and silhouette orientation with regard to vertical eye position. We discuss these results in the context of a simple image-based model to predict eye placement in arbitrary closed forms.

Acknowledgements: Funding from NSF BCS-2338600