Attention and automaticity in social judgments from facial appearance : cognitive and neural mechanisms

• Main Author: Raafat, R. M.
• Published: University College London (University of London) 2015
• Subjects: 150
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.668438

Recent evidence from behavioural and cognitive-neuroscience experiments has already yielded exciting discoveries into how we might code, process and perform judgments of facial social stimuli (indeed research into the latter provides a good vehicle for examining vision and object recognition in general). Nevertheless, the evidence regarding the role of top down control in face processing and the significance of the role of attentional capacity limits is contradictory or indeed absent in certain facial social trait judgments such as trustworthiness. In this thesis, I seek to present a portrayal of these roles, directed by load theory. Load theory suggests that perception has limited capacity but proceeds automatically on all stimuli until capacity is exhausted. Whether this process applies to arguably exceptional stimulus classes such as faces is contentious. Moreover, how this is related to social facial judgments such as trustworthiness, as compared to other evaluations such as threat and dominance judgments is unknown, as up until now, research on face-attention interactions has focused on directing visuospatial attention to emotional visual information rather than to facial trait judgments such as trustworthiness. In spite of this, the theory's predictions are clear; increasing the perceptual load of a task should consume capacity, thereby reducing processing of stimuli external to that task. Here I show that these predictions hold only for certain types of facial image evaluations but not for others. In a series of experiments that applied load theory, employing a combined visual search and face judgment task (where the level of attentional load in the search task was manipulated by varying the search set size of similar non-target letters), I find that under high perceptual load, observers become moderately less able to classify certain facial targets e.g. trustworthy ones as compared to dominant ones, even when these stimuli are fully expected and serve as targets. I also show the robustness of perceptual load effects by countering possible confounds and alternative explanations. Potential order effects are countered by reversing the order of the experiment, indicating that a possible attenuated short term memory imprint for the facial stimuli does not change the pattern of results previously experimentally demonstrated. Additionally, I find that high working memory load does not reduce social judgment evaluations under load, suggesting that perceptual biases during competitive interactions in visual processing are causative of the earlier demonstrated load effects. Following on from the modest but resilient results for trustworthiness modulation experimentally demonstrated here, the issue of bias for trustworthiness judgments is addressed in a signal detection paradigm (allowing bias to be discounted as a likely explanation of load effects). In the wake of the relatively robust results for trustworthiness perceptual load modulation, a new avenue for trustworthy judgments under attention is explored, investigating the possible role of dopamine in such evaluations in a clinical cohort of Parkinson's disease (PD) patients (as contrasted to age matched controls). PD has been linked with facial expression judgment impairment, although, this impairment could be subordinate to other cognitive processes enmeshed in facial evaluation, such as selective attention. Our results once again point to a pervasive role for perceptual load modulation of facial judgments, rather than a specific attentional deficit of PD. Finally, I explore the neurobiological correlates associated with facial social evaluations under perceptual load. In a neuroimaging study, I show neural responses to trustworthy facial images interact with attentional demands, demonstrating reduced activity under high perceptual load. I found high load only affects the facial components of trustworthiness (as compared to neutral faces) in cortical areas involved in social and facial processing (but not the facial signal components of untrustworthiness as compared to their neutral counterparts). The effects of load being specific to the trustworthy aspects of faces coheres with earlier presented behavioural results. As a final point, the demonstrated findings of negative-linear effects in the amygdala are consistent with prior research underlining the role of the amygdala in facial trustworthy judgments. This research presented here, although subtle in some experiments, provides convergent evidence that top-down cognitive and neural mechanisms are involved in influencing the degree to which facial visual judgments are processed. The results demonstrate the role of attentional modulation in facial social judgments and illuminate a possible role of perceptual load and attention in the facial automaticity debate. Both the type of facial judgment and category of facial valence are factors which determine the efficacy of perceptual load effects in facial evaluations.