Abstract: PURPOSE: Humans combine smooth pursuit and saccades in the ocular tracking of moving objects of interest. Although many studies have examined ocular tracking in children, these studies used predictive stimuli or stimuli of low uncertainty, thus smooth pursuit and saccades were assessed in conjunction with the predictive abilities. In addition, no study to date has examined the visual processing of target motion signals during ocular tracking in children. The current study aims to address these research gaps.
METHODS: We used an 8-minute ocular-tracking task based on the classic step-ramp paradigm modified to accommodate a full sampling of the polar angles. On each trial, participants tracked the step-ramp motion of a target (a cartoon character, 0.64°H × 0.64°V). Both the target speed and moving direction were randomly sampled from a range (speed range: 16-24°/s; direction range: 2-358° in 4° increment without replacement) to minimize expectation effects. A total of 78 children aged 8-9 years (female/male: 46/32) and 76 adults aged 18-30 years (female/male: 43/33) participated in this task. First, we computed 12 oculometric indices to measure different aspects of ocular-tracking performance and the dynamic visual processing of target motion. Second, previous studies have reported that open-loop pursuit relies on the reliability of visual processing of motion signals and can be affected by target motion signals in previous trials. In the current study, we thus examined the degree to which the open-loop pursuit response was affected by the target moving direction of the previous trials (i.e., the serial dependence effect). This helps reveal the robustness of visual processing of target motion signals in children.
RESULTS: All the 12 oculometric indices showed that the ocular tracking abilities in children were inferior to those in adults: (1) In the pursuit initiation stage when eye movements are primarily driven by input target motion signals (i.e., open-loop response), children had prolonged latency and slower eye acceleration; (2) in the steady-state tracking stage when eye movements are also driven by extra-retinal information (such as efference copy) about eye positions to correct tracking errors (i.e., closed-loop response), children’s pursuit velocity lagged more behind the target velocity, and the proportion of smooth pursuit was also lower; (3) for the saccadic eye movements, the frequency of saccades and the spatial distribution of saccade direction were similar in children and adults, whereas the amplitude of catch-up saccades was greater in children than in adults; (4) for tracking direction and speed, ocular tracking gains for open-loop pursuit direction and close-loop tracking speed were similar in children and adults but both were less precise in children than in adults. Of all the oculometric indices, the greatest difference between children and adults was in the latency of pursuit initiation (children’s mean at 0.3% of the adult population). In addition, for both children and adults, the pursuit direction in the open-loop response was pulled toward the target moving direction in the previous trial. Such serial dependence effect was stronger in children than in adults.
CONCLUSIONS: Our ocular-tracking task provides a wide range of largely independent oculometric indices that allow us to examine smooth pursuit and saccades and their coordination as well as visual processing of target motion signals in preadolescent children aged 8-9 years. Both open-loop and closed-loop tracking responses in children are inferior to those in adults. For the first time, we found that the gains of open-loop pursuit direction and close-loop tracking speed in children are comparable to those in adults, whereas the precision of both in children does not reach the adult level. This might be due to less reliable visual processing of target motion signals in children than in adults, which leads to the finding that the open-loop pursuit direction depends more on the recent history of target motion signals in children than in adults. We conclude that the development of different aspects of ocular tracking abilities follow different time courses, with the abilities related to open-loop pursuit maturing the last. The findings of the current study provide insights on the maturation process of cortical areas in charge of ocular tracking.

Key words: eye movements, smooth pursuit, saccades, motion perception, serial dependence