Disappearance of the inversion effect during memory-guided tracking of scrambled biological motion

Abstract

The human visual system is highly sensitive to biological motion. Even when a point-light walker is temporarily occluded from view by other objects, our eyes are still able to maintain tracking continuity. To investigate how the visual system establishes a correspondence between the biological-motion stimuli visible before and after the disruption, we used the occlusion paradigm with biological-motion stimuli that were intact or scrambled. The results showed that during visually guided tracking, both the observers' predicted times and predictive smooth pursuit were more accurate for upright biological motion (intact and scrambled) than for inverted biological motion. During memory-guided tracking, however, the processing advantage for upright as compared with inverted biological motion was not found in the scrambled condition, but in the intact condition only. This suggests that spatial location information alone is not sufficient to build and maintain the representational continuity of the biological motion across the occlusion, and that the object identity may act as an important information source in visual tracking. The inversion effect disappeared when the scrambled biological motion was occluded, which indicates that when biological motion is temporarily occluded and there is a complete absence of visual feedback signals, an oculomotor prediction is executed to maintain the tracking continuity, which is established not only by updating the target's spatial location, but also by the retrieval of identity information stored in long-term memory.

Global–local consistency benefits memory‐guided tracking of a moving target

“…Due to resource constraints and drop‐out rate of participants, however, we only recruited 22 right‐handed undergraduate or graduate students (50% female; mean age = 22.5 years, SD = 2.0) as subjects in this experiment and all participants received monetary compensation for it. This sample size was also similar to that used in most of the previous eye‐tracking experiments (Coppe et al., 2010; Jiang et al., 2016; Spering & Gegenfurtner, 2008). According to Daniël Lakens (https://psyarxiv.com/9d3yf/), if it is not possible to increase the sample size, the data analysis should not focus on p values, but on the effect size and the confidence interval, which are reported in Results of the present study to reflect the minimal statistically detectable effect.…”