McGrory, Catherine Ferguson
An investigation into the emotion-cognition interaction
and sub-clinical anxiety.
PhD thesis, University of Glasgow.
Full text available as:
This thesis combines behavioural and electrophysiological approaches in the study of the emotion-cognition interaction and sub-clinical anxiety. The research questions addressed in this thesis concern, specifically: the impact of emotion on attention; the interplay between attention and emotion in anxiety;and the cognitive construct of affect.
Chapter 1 provides an introduction to emotion research, cognitive models of anxiety and motivates the thesis.
Chapter 2 investigates whether affective processing is automatic. More specifically, to elucidate whether facilitated processing of threat in anxiety,
evidenced by emotion-related ERP modulations, requires attentional resources. It was previously reported that emotional expression effects on ERP waveforms
were completely eliminated when attention was directed away from emotional faces to other task-relevant locations (Eimer et al., 2003). However, Bishop et al. (2004) reported that threat-related stimuli can evoke amygdala activity without attentional engagement or conscious awareness in high-anxious but not low-anxious participants. Spatial attention was manipulated using a similar
paradigm as Vuilleumier et al. (2001) and Holmes et al. (2003), to investigate the mechanism underlying the threat-related processing bias in anxiety by examining the influence of spatial attention and trait anxiety levels on
established ERP modulations by emotional stimuli. Participants were instructed to match two peripheral faces or two peripheral Landolt squares. The Landolt squares task was selected since this is an attentionally demanding task and would likely consume most, if not all, attention resources. The ERP data did not offer support to the claim that affective stimuli are processed during unattended
conditions in high-anxious but not low-anxious participants. Rather, it questions whether a preattentive processing bias for emotional faces is specific to heightened anxiety. This is based on the finding of an enhanced LPP response for threat/happy versus neutral faces and an enhanced slow wave for threat versus neutral faces, neither modulated by the focus of attention for both high and low anxiety groups.
Chapter 3 investigated the delayed disengagement hypothesis proposed by Fox and colleagues (2001) as the mechanism underlying the threat-related attentional bias in anxiety. This was done by measuring N2pc and LRP latencies while participants performed an adapted version of the spatial cueing task.Stimuli consisted of a central affective image (either a face or IAPS picture, depending on condition) flanked to the left and right by a letter/number pair.
Participants had to direct their attention to the left or right of a central affective image to make an orientation judgement of the letter stimulus. It was hypothesised that if threat-related stimuli are able to prolong attentional
processing, N2pc onset should be delayed relative to the neutral condition. However, N2pc latency was not modulated by emotional valence of the central image, for either high or low anxiety groups. Thus, this finding does not provide
support for the locus of the threat-related bias to the disengage component of attention.
Chapter 4 further investigated the pattern of attentional deployment in the threat-related bias in anxiety. This was done by measuring task-switching ability between neutral and emotional tasks using an adapted version of Johnson’s (in press) attentional control capacity for emotional representations (ACCE) task. Participants performed either an emotional judgement or a neutral judgement task on a compound stimulus that consisted of an affective image (either happy versus fearful faces in the faces condition, or positive versus negative IAPS pictures in the IAPS condition) with a word located centrally across the image
(real word versus pseudo-word). Participants scoring higher in trait anxiety were faster to switch from a neutral to a threatening mental set. This improved ability to switch attention to the emotional judgement task when threatening
faces are presented is in accordance with a hypervigilance theory of anxiety. However, this processing bias for threat in anxiety was only apparent for emotional faces and not affective scenes, despite the fact that pictures depicting aversive threat scenes were used (e.g., violence, mutilation). This is discussed in more detail with respect to the social significance of salient stimuli.
Chapter 5 in a pair of experiments sought to investigate how affect is mentally represented and specifically questions whether affect is represented on the basis
of a conceptual metaphor linking direction and affect. The data suggest that the vertical position metaphor underlies our understanding of the relatively abstract concept of affect and is implicitly active, where positive equates with ‘upwards’ and negative with ‘downwards’. Metaphor-compatible directional movements were demonstrated to facilitate response latencies, such that participants were
relatively faster to make upward responses to positively-evaluated words and downward responses to negatively-evaluated words than to metaphorincompatible stimulus-response mappings. The finding suggests that popular use
of linguistic metaphors depicting spatial representation of affect may reflect our underlying cognitive construct of the abstract concept of valence.
Chapter 6 summarises the research in the thesis and implications of the present results are discussed, in particular in relation to cognitive models of anxiety.
Areas of possible future research are provided.
Actions (login required)