Children's visual guidance of self-selected stepping routes: Gaze-route correspondence in virtual reality

Children often negotiate physical challenges by selecting among several possible actions while coordinating vision and movement. This study examined how children's gaze corresponded with their self-selected stepping routes before and during locomotion and how gaze was directed towards the next foot placement. Children aged 7-10 years physically crossed a virtual river using a self-selected route across seven stepping stones while eye-tracking and foot-tracking data were recorded. The primary analytic sample comprised 350 children with valid task performance and sufficient eye-tracking data. Mixed-effects models compared sustained gaze visits to used and unused candidate stones across the pre-crossing and crossing phases. Gaze time was also examined among visually sampled stones, and transition-level analyses compared gaze towards the next selected stone with gaze towards other candidate stones. Before crossing, used stones were no more likely than unused stones to receive a gaze visit. During crossing, used stones were substantially more likely to be sampled. Among visually sampled stones, used stones also received considerably more gaze time during crossing. The next selected stone received a gaze visit in 98.1% of transitions and was markedly more likely to be sampled than other candidate stones. When visually sampled, the first gaze visit to the next selected stone began a median of 1.80 s before foot contact, and the upcoming stone received gaze during the final 500 ms in 90.4% of transitions. These findings show that children's gaze becomes strongly aligned with their self-selected route during locomotion and is prospectively organised around the next foot placement.