Global Precedence Effect
How do we process visual information? When we look at someone we know, we could recognize them by identifying individual parts of their face in isolation: the mouth, nose, ears, eyes, and son. But we could also recognize the configuration all the parts: how close their eyes are to their nose, the shape of the face and hairline, the height of their ears, and so on. Navon (1977) was interested in whether we process visual information piece-by-piece or in a more "global" holistic sense. To study that, he devised what are now called Navon hierarchical figures. Here's an example of a "Navon letter" in which a larger stimulus (a letter) is comprised of smaller stimuli (different letters):
{ width="100" }
The global precedence effect isn't universal. There are situations or tasks where the bias can be reduced, eliminated, or even reversed. For example, some recent research suggests the effect may be culturally specific. Navon's experiments were performed on a largely white, college-age population; since then investigators have found stronger global precedence effects in East Asian populations than were reported in Navon's study (McKone et al., 2010). Similar cultural differences in visual processing between North American and East Asian cultures have been demonstrated before (e.g., Kitayama, Duffy, Kawamura, & Larsen, 2003), and one theory is that the difference occurs because North American (specifically US) cultures are more individualistic, and East Asian cultures are more collectivist. Presumably because of this cultural difference, the visual system is tuned more to local (i.e., individual) features in the US and more to global (i.e., collective) features in East Asian cultures. Other researchers recently tested an African tribe and found that they demonstrated a local precedence effect (Davidoff, Fontenenau, & Fagot, 2008).
From Principal Components Enable A New Language of Images:
Human perception of visual stimuli has been shown to follow the global precedence effect [32], where the global information of the scene is processed before the local information. In [14], controlled experiments of presentation time on human perception of visual scene have further confirmed with the global precedence effect, where less information (presentation time) is needed to access the non-semantic, sensory-related information of the scene compared to the semantically meaningful, object- or scene-related information. Similar results have been reported in [3], where sensory attributes are more likely to be processed when the scene is blurred. Moreover, [33] have suggested that reliable structural information can be quickly extracted based on coarse spatial scale information. These results suggest that human perception of visual stimuli is hierarchical, where the global information of the scene is processed before the local information.