Holistic Face Processing is Penetrable…Depending on the Composite Design

Abstract

Holistic processing (HP) of faces, which is usually measured by the composite effect, may reflect an attentional strategy of processing all parts together. Two studies had evaluated this question previously. While Weston and Perfect (2005) found that priming at the local level speeded the recognition of the components of the faces and global priming did not have any effect on performance, Gao et al. (2011) found that only global priming had an effect on holistic processing of faces. The two studies used different versions of the composite task (the partial design, which is considered to be prone to biases, and the complete design) and also differed in other respects (e.g., 2AFC vs. same-different).Thus it is impossible to know to what extent issues with the partial design contributed to the different findings. To determine the consequences of priming local or global processing on HP we adopted the complete design measure of the composite effect and on each trial participants first attended either to the global or local level of Navon stimuli. As the partial design measure of the composite effect is nested within the complete measure, we were also able to obtain a partial design measure of HP and contrast the conclusions that would have been drawn from each design. The HP indexed by the complete design measure was augmented by global priming. In contrast, prior orientation of attention did not have any effect in the partial design index. We claim that the partial design index reflects other factors besides HP, including response bias, and conclude that the effects of global priming seem consistent with the hypothesis that HP can be understood within the context of domain-general attentional processes.