Abstract
Traditional approaches to the psychology of judgement and decision-making have made a great deal of progress in understanding the relationship between perception and action by constructing formal theories and instantiating them in computational models. The embodied cognition approach claims that studying only the reaction time and accuracy of decisions risks obscuring the dynamic nature of mental events. A confluence of these approaches has resulted in the rise of mouse tracking to investigate the dynamics of decision-making across many diverse sub-fields within psychology. However, models which link the process of decision-making to the recorded trajectories in mouse tracking experiments are rare, but frequently relied upon to lend mouse tracking research conceptual and theoretical support. Furthermore, there are more “embodied” ways to interact with the world, than pointing and clicking a mouse, such as reaching to grasp.This thesis aims to investigate whether the uncertainty introduced by perceptual decision-making influences the very ecological action of reaching to grasp an object, and whether this link can be modelled in a similar way to mouse trajectories. During my PhD I developed a novel experimental paradigm to use three-dimensional motion tracking to record reach-to-grasp movements in an experiment which balanced the rigor of traditional decision-making experiments with more everyday actions. These data were analysed using sophisticated statistical techniques including linear mixed modelling and distribution parameter fitting. An existing computational model which links noisy perceptual evidence accumulation to mouse path generation using embodied cognition principles was developed further to account for the additional complexity of three-dimensional reaches to grasp.
Overall, I found evidence that the reach-to-grasp actions took longer and were more curved when enacted with increased perceptual uncertainty. I concluded that these effects could not be entirely isolated to trials where there was a clear change of mind, but to a wider subset of reaches which were generated by a process distinct from the baseline trials. In addition, the computational model did not satisfactorily recreate the distributions of choice reach curvature generated by participants but indicated that there is a more complex relationship between decision dynamics and reach actions than is typically assumed by strong theories of embodied cognition.
Date of Award | 3 Oct 2023 |
---|---|
Original language | English |
Awarding Institution |
|
Supervisor | Iain D Gilchrist (Supervisor) & Casimir J H Ludwig (Supervisor) |