Abstract
Research using Garner’s selective attention paradigm suggests that, when we view static faces, the processing of facial identity interferes with the processing of expression, and vice versa (Ganel & Goshen-Gottstein, 2004). We recently replicated this result, but went on to show that interference is negligible when dynamic faces are viewed (Stoesz & Jakobson, submitted). This "dynamic advantage" could arise if, with the introduction of dynamic cues, viewers shift from using a global processing approach to focusing on local facial features (see Xiao et al., 2012). If this is the case, the advantage should be most apparent with upright stimuli, and in those with a global processing bias. To test these ideas, we assessed participants’ processing style using hierarchical stimuli, and then had them make speeded expression (or identity) judgements of static and dynamic faces presented in upright and inverted orientations while identity (or expression) was held constant (baseline block) or varied (orthogonal block). We calculated (a) corrected interference scores by determining the percent change from baseline RT seen in the orthogonal block; and (b) dynamic advantage scores by finding the difference between static and dynamic interference scores for each condition. As in our earlier work, interference was seen with static but not with dynamic stimuli, and the dynamic advantage was more evident with the expression than the identity task. However, planned comparisons revealed that the dynamic advantage seen during expression processing was eliminated after stimulus inversion for individuals showing a global (but not a local) processing bias. These results are consistent with the view that, when making expression judgments, global processors respond to the introduction of dynamic cues by switching to the use of a local processing strategy. Our findings highlight the importance of using dynamic displays and of considering individual differences when characterizing typical face processing mechanisms.
Meeting abstract presented at VSS 2013