Looking step by step

The general aim of this dissertation was to explore the extent to which gaze can provide insights into task performance in daily life. Previous research has often described gaze strategies as tightly coupled with task structure, particularly evident in sequential tasks like cooking or making tea. The expectation has been that once the sequence of actions in a task is understood, gaze behaviour may be easily predictable. However, this assumption might not hold true for tasks where central vision guidance is not continuously necessary, such as walking, driving, or stair climbing. The work presented in this dissertation focused on stair climbing. While vision is undoubtedly important for this task, it’s not always clear where one should look to perform it effectively. Laboratory studies have shown mixed results, with some suggesting a tight coupling between gaze and action, while others indicate that additional sources of information beyond central vision can guide our movements. In this dissertation, I examined gaze behaviour in familiar contexts comparing it with results coming from the lab, I explored factors influencing gaze beyond task structure, in particular task instructions. Finally, I focused my investigation on the exploration of the individual differences across participants. In Chapter 2, gaze behaviour during stair climbing in familiar settings was compared with lab studies, revealing high variability across participants and more flexible gaze patterns compared to the ones described in the lab. Chapter 3 investigated the effect of task instructions on gaze behaviour, showing that instructions explicitly mentioning stair climbing led to looking more at the staircase, whereas less explicit instructions resulted in more varied gaze behaviour away from the staircase towards the heading direction. In Chapter 4, the focus shifts to understanding how task difficulty affects gaze behaviour, finding that individual tendencies can play a more significant role than task difficulty. A differential reliance on peripheral vision was suggested as one of the potential factor guiding the different tendencies in directly looking at steps. Chapter 5 extended this investigation to blind individuals, showing similar variability in how they used their cane to touch steps, suggesting other sources of information besides peripheral vision, such as the reproducibility of the task and the expectation that all steps will be equal. The research highlighted that gaze behaviour is influenced by environmental context and task instructions, with significant differences across people. These findings suggest that gaze behaviour in daily activities, especially those not requiring constant central visual guidance, may not be tightly linked to task structure but rather to individual tendencies. This can have implications for rehabilitation strategies, which should consider individual differences in addition to group averages to optimise outcomes for patients with visual field defects.