Response inhibition in the parametric go/no-go task and its relation to impulsivity and subclinical psychopathy

ABSTRACT

The current study utilizes the parametric go/no-go task (PGNG), a task that examines changes in inhibitory performance as executive function load increases, to examine the link between psychopathic traits, impulsivity, and response inhibition in a cohort of healthy participants. The results show that as executive function load increased, inhibitory ability decreased. High scores on the Cognitive Complexity subscale of the Barratt Impulsivity Scale (BIS–11) predict poor inhibitory ability in the PGNG. Similarly, high scores on the Psychopathy Personality Inventory–Revised (PPI–R) Blame Externalization subscale predict response inhibition deficits in the PGNG, which loads more on the executive functions than the standard go/no-go task. The remaining BIS–11 as well as PPI–R subscales did not interact with inhibitory performance in the PGNG highlighting the specificity of associations between aspects of personality and impulsivity with inhibitory performance as cognitive load is increased. These data point towards the sensitivity of the PGNG in studying response inhibition in the context of highly impulsive populations and its utility as a measure of impulsivity.

KEYWORDS:

• Parametric go/no-go task
• Impulsivity
• Psychopathy
• Barratt Impulsiveness Scale
• Psychopathic Personality Inventory–Revised

Acknowledgement

The authors would like to thank C. V. Dolan, Department of Biological Psychology, VU University, Amsterdam, the Netherlands, for help with the manuscript.

Notes

¹Although not directly relevant to the argument here, it should be noted that in addition to the ongoing debate about what should be considered an EF, there have been arguments concerning whether the suite of EFs constitute a unitary construct or a number of diverse functions (e.g., Baddeley, 1996; Duncan, Johnson, Swales & Freer, 1997; Miyake et al., 2001).

²The data presented here were collected in two rounds. The first round of data collection involved 60 participants. Based on an interim analysis of the first 60 participants, data from a further 26 participants were deemed necessary to achieve the required power. To alleviate any concern regarding a potential inflation of the statistical outcome, theoretically critical findings were also analysed using a Bayesian approach where additional data collect/analyse cycles do not carry the same potential for error inflation as frequentist statistics.

³The PPI–R factor Impulsive Antisociality expressed the same relationship to response inhibition as Blame Externalization [F(1, 84) = 6.03, MSE = 281.85, p_uncorr = .02, , CI = [0.004, 0.11]; R² = .07, b = −0.15], but this finding did not survive correction for multiple comparisons. The analyses on Blame Externalization as well as Cognitive Complexity were additionally carried out with repeated measures ANCOVAs, adding gender as a covariate. The results showed no influence of gender (all Fs > .6), and the inclusion of gender as a covariate did not alter the significance of the remaining results on Blame Externalization and Cognitive Complexity.

⁴To note, the effect sizes reported here are of the same magnitude as those found in Keilp et al. (2005), where the standard version of the go/no-go was employed (r = .37–.38). The effect size obtained here is comparable to that in previous research on the relationship between BIS–11 impulsivity and other versions of the go/no-go task. However, Keilp et al. (2005), examined the relationship between the Attentional and Motor Impulsiveness factor scores and go/no-go performance, and as such it is unclear which individual BIS–11 subscales were related to performance in the standard go/no-go task, whereas in the PGNG a specific relationship between response inhibition accuracy and impulsivity relating to Cognitive Complexity existed.