Perspective taking and systematic biases in object location memory

Abstract

The aim of the current study was to develop a novel task that allows for the quick assessment of spatial memory precision with minimal technical and training requirements. In this task, participants memorized the position of an object in a virtual room and then judged from a different perspective, whether the object has moved to the left or to the right. Results revealed that participants exhibited a systematic bias in their responses that we termed the reversed congruency effect. Specifically, they performed worse when the camera and the object moved in the same direction than when they moved in opposite directions. Notably, participants responded correctly in almost 100% of the incongruent trials, regardless of the distance by which the object was displaced. In Experiment 2, we showed that this effect cannot be explained by the movement of the object on the screen, but that it relates to the perspective shift and the movement of the object in the virtual world. We also showed that the presence of additional objects in the environment reduces the reversed congruency effect such that it no longer predicts performance. In Experiment 3, we showed that the reversed congruency effect is greater in older adults, suggesting that the quality of spatial memory and perspective-taking abilities are critical. Overall, our results suggest that this effect is driven by difficulties in the precise encoding of object locations in the environment and in understanding how perspective shifts affect the projected positions of the objects in the two-dimensional image.

Keywords: 3D perception: Space Perception; Spatial cognition; Spatial memory.

Fig. 1

a Top-down schematic of the virtual environment used in the experiment with camera positions. Green cameras represent camera positions at encoding, and blue cameras represent the corresponding camera positions at test. Red arrows indicate the direction of perspective shift for each of the encoding cameras. Example renderings from the encoding (Start Position 1) and test camera are also provided. b Trial structure. (Color figure online)

Fig. 2

The upper panel shows example stimuli from the learning and the test phases for congruent and incongruent trials across 13 cm and 61 cm object movements. The bottom panel shows accuracy as a function of distance (cm) and congruency, with mean (solid line) and 95% CIs (gray shaded area) and individual data points and violin plots behind. Chance level performance is indicated by the solid horizontal line

Fig. 3

Stimuli examples for the no objects and additional objects in incongruent and congruent trials for a 13 and 61 ODD

Fig. 4

Bar plots for accuracy values as a function of congruency (incongruent/congruent) and condition (no objects/additional objects) and age group, with mean (solid line) and 95% CIs (gray shaded area) with individual data points and violin plots behind

Fig. 5

Bar plots for the accuracy as a function age group (young/older adults) and congruency (congruent/incongruent), with mean (solid line) and 95% CIs (gray shaded area) with individual data points. The solid horizontal black line indicates chance level performance

Fig. 6

a Bar plots for the accuracy for each cluster as a function of congruency and ODD, with mean (solid line) and 95% CIs (gray shaded area) of individual data points. b Mean accuracy in each cluster in congruent and incongruent trials with individual data points