Electrophysiological Correlates of Intellectual and Emotional Intelligence
Burns, Christopher G.
This thesis examines the electrophysiology of several inspection time (IT) tasks, specifically extending two strands of work from Edinburgh by Austin (2004 and 2005) and Zhang et al. (1989a and 1989b). Austin designed a new emotional inspection-time task using human faces as stimulus items. The original pi-figure IT task, extensively investigated since 1970, has been found to generate robust correlations with assorted IQ measures. A potential confound in existing IT methodologies is that the IT-IQ relationship may not arise from particular stimulus presentation methodologies, but be due to a process of rapid strategy formulation. Variation in the stimulus forms (e.g. pifigures, human faces, geometric shapes or auditory tones) affects the robustness of the IT-IQ relationship. Austin’s tasks were modified to permit the acquisition of ERP data to examine the effect of stimulus emotional expression and to explore the relationship with existing psychometric scales. Early differences in ERP related to IQ were reported by Zhang. A key additional element of this thesis is the examination of relationships between ERP and the emotion shown on stimulus faces, since differences in emotional expression form the basis of the emotional-IT task. Four major experiments were conducted. Experiment 1 piloted the face presentation task and set baseline timing values for presenting human face stimuli; participants identified gender from each stimulus. A psychophysical curve was constructed, and the difficulty scaling of the task suggested that the longer timing durations should be removed in favour of additional, shorter durations in subsequent experiments. Experiment 2 was the first attempt in the literature to collect ERP data from the emotional-IT task. The expected negative correlations between psychometric IT and IQ measures were reproduced, but the correlation between IT and EI scores was found to be positive; higher EI scores resulted in slower IT values. A P100 -> N170 -> P300 ERP complex was evoked, with maximal amplitudes at parietal electrode sites, and maximal activations in response to happy-face stimuli, especially among males. When divided into high- and low-IQ groups, higher-IQ individuals showed steeper mean gradients, and gradient-IQ correlations 50ms earlier than among the lower-IQ group. Experiment 3 evaluated different backward masking techniques. In the ERP data, traces elicited by the stimulus face could potentially be contaminated by activity related to the backward mask. A P100®N170®P300 ERP complex was reproduced, but despite very high participant success rates (95.3%), effects of stimulus emotion within this ERP were not pronounced. A newer non-face mask was adopted for future studies to minimise other mask-contamination confounds from larger population samples in subsequent experiments, and to avoid potential apparent motion effects, another known confound in IT methodologies. Experiment 4 featured three consecutive ERP acquisitions (face-IT-1, line-IT and face-IT-2) and was analysed in two stages. ERP effects related to stimulus emotion were inconsistent; the responses to stimulus emotion were neither identical nor prominent in each emotional-IT task. Psychometric effects were more consistent. IQ and IT were negatively correlated as expected, while IQ, IT and emotional intelligence were positively correlated. Throughout the present series of experiments, the expected relationships between IT and IQ were robust across non-traditional emotional-IT tasks. The effects of stimulus emotion on ERP traces were not prominent despite relatively large sample sizes and adequate effect-size estimates. The ERP relationships with IQ previously found at Edinburgh and by others in line-IT tasks were not replicated here, although the lack of such a relationship has precedence in the broader literature.