Seminar Series

Age-related and individual differences in the first half-second of visual processing

Human beings are both very fast and accurate at detecting objects in natural scenes, for instance animals and vehicles (http://journalofvision.org/3/6/5/; http://www.journalofvision.org/ 4/1/2). Many papers have claimed that our visual system is even faster at processing one category of objects: faces. However, measuring visual processing speed is challenging because of [1] uncontrolled physical differences that co-vary with object categories; [2] the effect of task constraints on diagnostic information; [3] statistical issues. Controlling the stimulus space allow us to provide better estimates of the time course of face processing. For instance, early ERP differences between faces and objects are preserved even when images differ only in phase information, and amplitude spectra are equated across image categories (http://journalofvision.org/8/12/3/). Beyond simple categorical designs, parametric designs can be used to establish the relationship between stimulus space and brain activity. A single-trial GLM approach can help us compute more meaningful averages of EEG/MEG data (http://www.biomedcentral.com/1471-2202/9/98). This approach can be used for instance to study visual processing speed in younger and older observers (http://frontiersin.org/perception%20science/10.3389/fpsyg.2010.00019/abstract; http://www.biomedcentral.com/1471-2202/1410/1114), as well as task effects on face processing. In this context, I will present data suggesting that group statistics can be misleading and hide unexplained inter-subject variability.