Abstract
Psychopathy is a complex, newsworthy psychological construct: a deviation from normal functioning that impacts an entire personality spectrum. Current neurobiological and neurocognitive models attempt to explain its underlying neural mechanisms using a top-down approach. But none account for all facets of psychopathy. This thesis uses a bottom-up approach, testing whether the existing neuropsychologically-based Reinforcement Sensitivity Theory (RST) can explain psychopathy traits. RST includes three conceptual systems: Goal Attraction System (GAS), Goal Repulsion System (GRS), and Goal Inhibition System (GIS). These have been applied to psychopathy previously but have not used validated neural measures. Demonstration of a neural basis based on RST could indicate potential pharmaceutical treatments.
Our initial analysis (Chapter 4) used an archival dataset to investigate the relationship between psychopathy and neural RST constructs. We found small correlations between psychopathy and these RST constructs, which were published in Personality and Mental Health. Subsequently, we used a new sample of 250 individuals from the Dunedin community, overweighted for psychopathy symptoms, for replication and extension. Participants completed self-report measures, then behavioural tasks while connected to an electroencephalogram (EEG), and finally underwent a semi-structured interview. Within the battery of measures were two self-report psychopathy scales and two clinician-rated psychopathy measures. The tasks provided a range of EEG measures of each of the RST systems.
Chapter 5 reports a factor analysis of all psychopathy measures, resulting in a four-factor model: boldness, disinhibition, affective, and interpersonal. Chapter 6 reports the attempt to generalise and replicate the initial analysis findings. The results were unexpected, so I conducted post-hoc analyses of variance to investigate non-linear relationships between psychopathy domains and neural RST constructs. Chapter 7 reports an extension of these findings using a different EEG task. Only the GIS measure was associated with partial validation as an RST construct measure; and so only it was used to examine non-linear relationships with psychopathy domains.
Overall, most findings were null or small, with some unexpected results. They did not provide a comprehensive view of how RST systems work together to explain psychopathy. Findings for specific psychopathy domains were: 1) Boldness was unexpectedly associated with higher levels of GIS (i.e., anxiety), contradicting existing literature. 2) Disinhibition traits might be explained by increased sensitivity in the GIS (i.e., anxiety) and an unexpected negativerelationship with the GAS (i.e., approach) was found. 3) Affective traits might be explained by decreased sensitivity in the GIS and GRS (i.e., fear), but an unexpected positive association was also found in the alpha frequency range for the GRS. I also investigated relationships between psychopathy and self-report scales of RST throughout the two samples, with findings as expected and aligning with past literature. These findings make it clear that the self-report and neural measures of RST are measuring different constructs.
While these findings imply that RST may not be the most suitable framework to understand psychopathy comprehensively, some relationships were identified between psychopathy domains and RST systems, particularly with the GIS. Considering this, tentative implications are discussed.