Dominique Lopiccolo¹ (), Alon Hafri¹; ¹University of Delaware

Representational momentum (RM) is a bias in visual memory whereby moving or changing objects are remembered as farther along their trajectory than they actually were. Most investigations of RM have involved rigid spatial transformations (e.g., rotation), but less is known about RM for nonrigid transformations. Finding RM in this context would expand our understanding of the extent and limits of anticipated change that are encoded in memory. In earlier work, we examined RM for spatial scaling, a nonrigid change in which an object resizes while maintaining its shape. Dynamic displays of expanding or contracting circles elicited RM, such that the circles were remembered as larger or smaller than their final size. Yet a striking asymmetry emerged: RM appeared consistently greater for shrinking than for growing. Here we asked whether this asymmetry reflects an interaction with an independent “smaller bias” previously observed in visual memory, whereby static objects are remembered as smaller than they appeared. If such a bias exists, it would align with shrinking events, increasing the propensity to report the later (smaller) frame, but oppose growing events, decreasing that propensity. To test this, participants viewed two trial types: in dynamic trials, circles grew or shrank before being masked; in static trials, they remained stationary. Participants then completed a two-alternative forced-choice task, selecting which of two probes (one smaller, one larger) matched the last image they saw. We again observed a robust RM effect and reproduced the shrink–grow asymmetry. Crucially, static trials also revealed a reliable smaller bias, and after accounting for this bias, RM magnitude was equal across scaling directions. Together, these results show that the apparent asymmetry between growing and shrinking RM arises from the interaction of two distinct memory biases. More broadly, they reveal how multiple mechanisms jointly shape visual memory, even for simple geometric transformations.