The role of memory and perspective shifts in systematic biases during object location estimation

Abstract

In the current study, we investigated whether the introduction of perspective shifts in a spatial memory task results in systematic biases in object location estimations. To do so, we asked participants to first encode the position of an object in a virtual room and then to report its position from memory or perception following a perspective shift. Overall, our results showed that participants made systematic errors in estimating object positions in the same direction as the perspective shift. Notably, this bias was present in both memory and perception conditions. We propose that the observed systematic bias was driven by difficulties in understanding the perspective shifts that led participants to use an egocentric representation of object positions as an anchor when estimating the object location following a perspective shift.

Biases in object location estimation: The role of rotations and translation

“…In line with this explanation, previous research suggests that adjustments require time and cognitive effort (Epley et al, 2004 ) and as a result, individuals often stop adjusting their responses once a plausible estimate is reached thereby biasing their responses in the direction of the initial anchor (Quattrone, 1982 ; Tversky & Kahneman, 1974 ). If participants in our previous studies (Segen et al, 2022 ; Segen et al, 2021a ) also showed insufficient adjustments, this would explain the systematic bias in the direction of the perspective shift.…”

“…Our conjecture is that the perspective shift-related bias is caused by egocentric influences on target object estimates (Segen et al, 2022 ; Segen et al, 2021b ). If participants relied solely on an allocentric representation in which the object position is encoded relative to other features in the environment, their own position and movement in the environment should not influence their responses.…”

“…In our previous work (Segen et al, 2022 ; Segen et al, 2021a ; Segen et al, 2021b ), participants may have used the egocentric self-to-object vector as an anchor for their response, which would result in participants “dragging” the object with them following a perspective shift. In line with this explanation, previous research suggests that adjustments require time and cognitive effort (Epley et al, 2004 ) and as a result, individuals often stop adjusting their responses once a plausible estimate is reached thereby biasing their responses in the direction of the initial anchor (Quattrone, 1982 ; Tversky & Kahneman, 1974 ).…”

“…Our recent research suggests that such paradigms may yield a systematic bias in reporting memorized object locations (Segen et al, 2022 ; Segen, Avraamides, Slattery, Colombo, & Wiener, 2021a ; Segen, Colombo, Avraamides, Slattery, & Wiener, 2021b ). Specifically, we found that when participants were asked to indicate the location of an object following a perspective shift (Segen et al, 2022 ; Segen et al, 2021a ) or when asked to judge the direction in which an object moved after a perspective shift (Segen et al, 2021b ), they systematically localized the object further in the direction of the perspective shift. Interestingly, this systematic bias in the direction of the perspective shift is not driven by distortions introduced in memory as participants also exhibited this bias in a perceptual version of the task (Segen et al, 2022 ).…”

“…It is also possible that the uncertainty about the perspective shift arises due to the way we introduced perspective shifts. For example, in our previous studies, the perspective shift consisted of a combination of camera translation and rotation movements (Segen et al, 2022 ; Segen et al, 2021a , 2021b ). Specifically, the camera moved in a circle such that it translated in one direction and at the same time rotated in the opposite direction.…”