Degree Name

Doctor of Philosophy


Department of Psychology


Studies on the global precedence effect have typically been concerned with the relative timing and interference of the global and local levels of a compound pattern. Navon (1977) purposed that global configurations are processed more quickly than are local components. Kinchla and Wolfe (1979) demonstrated that this global precedence was lost when the global stimuli exceeded 6-9 degrees of visual angle. Navon and Norman (1983) noted that Kinchla and Wolfe confounded size and eccentricity. They attempted to unconfound these two variables and concluded that global precedence was found with both large and small stimuU when eccentricity was not a confound. It is argued that the controversy in the global precedence hypothesis has partially resulted from the fact that the concepts of the global and local and the attributed relationship between these two vary in different experiments and among workers. In the present study a new approach to global-local relationships was introduced. Global-local relationships were classified into four groups: Inclusion, placeholder, dimensional, and featural relationships. Any of these relationships may or may not be accompanied by a semantic relationship between the global and local levels. The Stroop-type interference of the global and local levels refers to some sort of conventional semantic relationship between the levels. Based on an extensive study of the literature it was suggested that the perception of the global and local levels of a compound pattern is dependent on the type of relationship between two levels. The presented study was limited to test the effect of size and eccentricity of the compound pattern on the processing of the global and local levels with inclusion and placeholder relationships.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.