Correlates of novel Bayesian Estimation of response inhibition parameters

Response inhibition is a core executive function, used to cancel or withhold behaviour and can be operationalised using the stop-signal task. The main dependent variable of this task is stop-signal reaction time (SSRT), an estimate of the latency of the inhibitory process. While SSRT has been shown to be deficient in a number of impulse control disorders and psychiatric conditions, it cannot account for the full range of processes necessary for effective response inhibition. This thesis challenges the notion that response inhibition is completely captured by the single summary measure of SSRT and instead suggests that reliable triggering of the inhibitory process (i.e., trigger failure) is as important, if not more so, than its latency. A Bayesian ex-Gaussian hierarchical model of inhibitory control was used, which estimates both SSRT and trigger failure. Differences between traditional and model estimations of SSRT were undertaken alongside correlational approaches to explore the impact of the novel estimation technique on the conceptualisation of response inhibition. The inclusion of trigger failure both improved the fit of the data and substantially attenuating estimates of SSRT. This attenuation drove changes to the relationships between SSRT and both impulsivity and electrophysiology. Correlations between impulsivity and SSRT were lost after accounting for trigger failure. Likewise, electrophysiological markers thought to represent response inhibition were not found to relate to model estimated SSRT. Instead attentionally sensitive electrophysiological components were related to both SSRT and trigger failure. This work shows that response inhibition is a multifaceted process, which cannot be characterised by a measure of inhibition latency alone. Instead it relies on a combination of attention to the stop-signal that appropriately triggers a fast inhibitory process. This reconcepualisation of the processes involved in response inhibition has implications for both past and future research aimed at understanding the control of behaviour.