Date of Completion


Embargo Period



visual perception, recognition, point-light display, multifractal

Major Advisor

James A. Dixon

Associate Advisor

Claudia Carello

Associate Advisor

Claire F. Michaels

Associate Advisor

Jeffrey-Kinsella Shaw

Field of Study



Doctor of Philosophy

Open Access

Open Access


This dissertation investigates the potential contribution of fractal fluctuations of head sway in the time evolution of visual recognition in biological motion perception. The first experiment found no difference in recognition times when point light display (PLD) activities are shown either from a fixed or a moving point of observation. The second experiment, using head tracking, multifractal analyses, and geometrical manipulations in the PLDs found that (1) the multi-scale fractality of head sway is different before and after recognition, and (2) the time-evolution of the multifractal spectra predicts recognition. The third experiment manipulated both the geometrical qualities of PLDs and the context of the presentation (e.g., showing the PLDs multiple times, priming the PLDs with congruent or incongruent images of objects, and adding contextual changes). In all cases, the time-varying multifractal structure of head sway predicted changes in visual perception, in particular, the transition from exploration to recognition. Event history analyses indicated a reliable contribution of the width of the multifractal spectrum width to the evolution of recognition. Despite a wide range of geometrical and contextual manipulations on the PLDs, multi-scale interactions remained a strong index of the self-organization of biological motion perception.