Viewpoint dependence in scene recognition: Testing the spatial limits of the Reaction Time-based Concealed Information Test
Аннотация
Background The Reaction Time-based Concealed Information Test (RT-CIT) is a highly effective memory detection paradigm that relies on the congruency between a suspect’s internal memory and external test stimuli. While previous research indicates the RT-CIT is remarkably resilient to angular deviations when testing isolated objects, its efficacy and boundary conditions for complex, three-dimensional spatial scenes remain unexplored. Objective The present study investigated whether the robustness of the RT-CIT extends to spatial environments by testing how angular shifts between encoded and tested viewpoints impact memory detection. Method Participants encoded 10 unique interior scenes (5 Targets and 5 Probes) from a single viewpoint. During the RT-CIT phase, participants were tested on these scenes rendered at five specific angular deviations: 10°, 45°, 90°, 135°, and 170°. Results The RT-CIT effect (the Probe-Irrelevant RT difference) remained robust across moderate viewpoint shifts, maintaining significance up to a 125° deviation. The RT-CIT effect at the 160° deviation did not reach statistical significance. However, we still observed a small-to-moderate effect size ( d = 0.371), possibly suggesting an attenuated rather than absent effect. Furthermore, unlike previous findings with isolated objects, spatial scenes exhibited an immunity to absolute presentation angle effects. The RT-CIT effect also maintained high temporal stability across all test epochs. Conclusion These findings establish the boundary conditions of the RT-CIT: While global layout features provide robust memory anchoring against moderate rotations, extreme perspective inversions possibly attenuate CIT efficacy.