The Face Inversion Effect: Investigating the role of perceptual learning, facial specificity, and holistic processing.

Abstract

This thesis investigates the perceptual processes underlying face recognition and the face inversion effect, exploring whether there is evidence for facial specificity in this effect and the specific types of information that produce it. Results will be presented from behavioural studies using a variety of manipulated face stimuli and studies involving transcranial direct current stimulation (tDCS) using face stimuli and prototype-defined checkerboard stimuli, which have previously been used to demonstrate the role of perceptual expertise in the face inversion effect (McLaren, 1997; Civile, Zhao, et al., 2014; Civile, Verbruggen, et al., 2016). Chapter 1 outlines the previous literature and background theories underpinning the face recognition research. Chapter 2 directly compares the effect of tDCS on the inversion effect for faces and checkerboards and tests a new active control condition. The disparity in the remaining inversion effect for faces and checkerboards under tDCS has led to the suggestion that there may be an additional, potentially face-specific component contributing to the inversion effect for faces together with perceptual learning. The findings here offer some support for this idea and also indicate based on the active control comparison that it is the specific Fp3-Fp2 montage that produces this tDCS induced effect. Chapter 3 begins a series of experiments exploring the idea that holistic processing (indexed by face contour) may be part of this additional component. Scrambled faces were used on the basis that they have been shown to result in a robust inversion effect despite complete disruption to configural information (Civile et al., 2014), and were subject to a contour manipulation to assess the impact of this on the inversion effect. Results showed that disruption to the contour information in addition to scrambling was sufficient to reduce the inversion effect. Chapters 4 and 5 extend this contour manipulation to normal faces and New Thatcherised faces to further explore the impact on the inversion effect. Results from these studies are somewhat mixed with some showing that contour manipulation reduces overall performance only, and others indicating that it impacts the inversion effect. Chapter 6 aims to investigate how tDCS stimulation is able to produce the effect on face recognition that it has been shown to and does so by utilising the typical anodal Fp3-Fp2 montage and then reversing the polarity to deliver cathodal stimulation. This reversal was shown to also reverse the behavioural effects, with anodal stimulation resulting in a reduction to the inversion effect and subsequently delivered cathodal stimulation increasing it again. Chapter 7 summarises the experimental findings and discusses the implications in terms of the wider literature as well as offering suggestions for future research.