Abstract
Holistic processing (HP) is widely believed to play a critical role in face processing and recognition, and a variety of tasks have been created to gauge the respective contribution of HP and non-HP (analytic) in faces. However, there has been some controversy regarding the role of HP in face memory with respect to individual differences. One possible reason for the controversy may have to with the different approaches used to measuring HP. In the present study, we first compared the subtraction and regression approaches to measuring HP in three standard face-processing tasks (component, configural, and composite), and found that, compared to subtraction, residues after regressing the non-HP element from each face processing task can provide purer measures of HP. Moreover, the HP estimate of the component task can best predict performance on the Taiwanese Face Memory Test (TMFT). We next investigated whether neural activity in the face-selective brain regions, including fusiform face area (FFA), occipital face area (OFA), posterior region of superior temporal sulcus (pSTS), and ventral anterior temporal lobes (vATLs), can effectively predict HP performance on the three face processing tasks. To that end, we first identified during the functional scan those regions by asking participants to perform a one-back task, while viewing either static images or dynamic videos. We then determined for each region of interest the cluster size associated with maximum face selectivity. Finally, correlation analyses revealed that participants with greater BOLD signals in FFA and vATL demonstrated better performance on the HP of face processing tasks. Taken together, our findings indicate that (a) the regression approach can provide more robust measures of HP, (b) HP of the component task can best predict performance on face memory, and most importantly (c) connections between regression-based measures of HP with brain regions selective for face processing and face memory.
Meeting abstract presented at VSS 2017