Date created: | Last Updated:

Category: Hypothesis

Description: We lack unified cognitive measures of impulsivity that assess the different processes linked to impulsive behaviours while controlling for lower-level perceptual and motor demands. We developed the Cognitive Impulsivity Suite (CIS), which measures attentional focus, information gathering and monitoring / shifting aspects of impulsivity using a harmonised set of gamified tasks (https://www.youtube.com/playlist?list=PLs-GMH-Foyaz-UkhauX-uSCDBQkAM1sTv). We conducted an initial validation study of the CIS in 1,056 healthy participants tested online via Amazon Mechanical Turk and 63 healthy participants tested in face-to-face laboratory sessions (preprint: https://psyarxiv.com/bynsm/). Findings supported the three-factor structure of the CIS (Attentional Control, Information Gathering and Monitoring / Shifting) and provided evidence for its convergent and criterion-related validity (assessed against the SUPPS-P impulsivity scale, and alcohol, drug and gambling use identification scales, i.e. AUDIT, DUDIT and PGSI respectively). Here, we aim to confirm the factor structure of the CIS and its convergent and criterion validity, using the same validation measures (SUPPS-P, AUDIT, DUDIT and PGSI) plus an ADHD screening (as an additional index of ‘real-world’ impulsivity) in an independent sample comprising community-dwelling participants and a patient population of alcohol, drug and gambling users engaged in online counselling treatment. HYPOTHESES Hypothesis 1: Factor mixture models that represent the community-based and clinical samples as a single homogenous population in which participants vary in severity along latent dimensions of cognitive and self-reported impulsivity, alcohol use, drug use, gambling problems and symptoms of ADHD, will provide a better fit to the data than an alternative model that distinguishes between these sample as distinct clinical and non-clinical groups. Hypothesis 2: A three-factor model differentiating between Attentional Control, Information Gathering, and Monitoring / Shifting as empirically distinct factors will provide a superior fit to performance data obtained on the CIS compared to competing one- and two-factor models. Hypothesis 3: The three CIS factors will exhibit different patterns of intercorrelations with the five facets of trait impulsivity measured by the SUPPS-P: Information Gathering will exhibit stronger negative correlations with Positive Urgency and Sensation Seeking compared to the other facets of trait impulsivity; Monitoring / Shifting will be more strongly negatively correlated with Positive Urgency, Negative Urgency, and Lack of Premeditation compared to the other facets of trait impulsivity; and Attentional Control will exhibit stronger negative correlations with Positive Urgency compared to the other facets of trait impulsivity. Hypothesis 4: The CIS factors will demonstrate incremental criterion-related validity in explaining a meaningful proportion of variance in real-world addiction problems above that explained concurrently by self-reported impulsivity as measured by the five SUPPS-P subscales. We expect a negative linear relationship of at least small effect between Monitoring / Shifting and the addiction problems factor (with AUDIT / DUDIT / PGSI as factor indicators). We also expect the relationships will be stronger than those reported in the original validation of the CIS in a community sample.