Boundary extension is constrained by naturalistic image properties
Boundary extension (BE) is a classical memory illusion in which observers remember more of a scene than was presented. According to predictive accounts, BE reflects the integration of visual input and expectations of what is beyond the boundaries of a scene. Alternatively, according to normalization accounts, BE reflects one end of a normalization process towards the typically-experienced viewing distance of a scene, such that BE and boundary contraction (BC) are equally common. Here, we show that BE and BC depend on depth-of-field (DOF), as determined by the aperture settings on a camera. Photographs with naturalistic DOF led to the strongest BE across a large stimulus set, while BC was primarily observed for unnaturalistic DOF. The relationship between DOF and BE was confirmed in three controlled experiments that isolated DOF from co-varying factors. In line with predictive accounts, we propose that BE is strongest for scene images that resemble day-to-day visual experience.