Abstract
Natural scenery consists of objects built from self-repeating patterns, including clouds, mountains, and trees. Mandelbrot introduced the term “fractal” in 1975 to describe this family of objects. We rarely perceive one independent stimulus at a time in the natural world. Instead, we more often process layers of stimuli. Therefore, natural stimuli are better represented through layered fractal images. The current research behaviorally evaluates how we process layers of fractal information, otherwise known as composite fractals. The study utilized a within subjects design to examine the relationship between visual complexity, measured with fractal dimension (D), fractal type, and preference. Two types of fractal stimuli were created: a “mountain” fractal and a “cloud” fractal (see Bies et al., Symmetry, 2016). The experiment consisted of four blocks, with the first three using two-alternative forced-choice tasks in which pairs of fractals patterns of differing complexity were presented simultaneously and participants selected their preferred pattern. Block one paired mountain fractals, block two cloud fractals, and block three composite fractal stimuli consisting of a combination of a mountain fractal placed below and overlapping a cloud fractal of the same level of complexity. Block four had participants rate their preference for composite fractal stimuli with varying fractal dimension. The results indicated the highest preference for mountain fractals of D = 1.1, and cloud fractals of D = 1.7. For composite fractals with the same D, preference peaked at D = 1.5. The rating portion of the study revealed that there is significant preference for composite fractals where the mountain fractal is of lower complexity than the cloud fractal. We discuss the results within a fractal fluency model in which people have acquired their preference through exposure to prevalent fractal landscapes.