How does the mind coordinate local and global features of a display to allow for adaptive functioning? To answer this question, we presented adults with a speeded categorization task in which they had to decide whether two stimuli match in a local element, in their global pattern, or in neither the local nor the global feature. The trial series of reaction times were then subjected to fractal analyses to capture the coordination that gives rise to performance. The assumption is that long-range correlations reveal themselves in pink-noise exponents, ones that are higher than white-noises exponents. To investigate the stability of fractal exponent, we manipulated both the local elements (to be either familiar or novel), and the order of trials (to be either random or blocked). Results show a significant deviation from white-noise, but only in familiar-elements condition in which trials were presented randomly. Implications for local/global research are discussed.