A simple theoretical agent-based model intending to represent anxiety in students is presented. In this, agents make two decisions each simulation tick, whether to: socialise and go to school (on school days). Circular cau-sation can result in some agents entering a negative loop of high-anxiety and avoidance, even whilst most improve their skills and develop lower levels of anxiety. We compare three different mechanisms that may lie be-hind decision-making under anxiety: expect the worst, avoid uncertainty and avoid anxiety. Results indicate that, although these are often indistin-guishable in terms of some obvious aggregate measures on the outcomes, they do exhibit different underlying dynamics.

This study describes the calibration of a dynamic simulation model of an individual 'suicidal mind' using ecological momentary assessment (EMA) data collected from 115 military veterans with PTSD, approximately 50% of whom reported suicidal ideation at baseline. The model synthesizes the Integrated Motivational-Volitional (IMV) model of suicide, the Fluid Vulnerability Theory, and the Cusp-Catastrophe model into a system-dynamics framework that estimates an individual's level of suicidality over time. In prior work, the model was validated across 4.2 million simulations using arbitrary parameter values, successfully reproducing theoretical characteristic pathways to suicidal behavior proposed in the literature. This extension grounds the model in empirical data by calibrating individual-level parameters to high-frequency EMA time series, capturing suicidal ideation, affect, stress, and alcohol use across four daily prompts over two weeks, using genetic optimization. The calibrated model is then used to evaluate whether simulated clinical interventions, including PTSD therapy and crisis response planning, produce trajectories consistent with observed treatment outcomes. This work-in-progress demonstrates a methodology for bridging theory-driven simulation and empirical clinical data, with implications for virtual case studies, personalized treatment evaluation, and the refinement of psychological theories of suicidal behavior.

Shared understanding among stakeholders is widely recognized as a central condition for successful collaborative governance, yet the mechanisms through which it emerges remain poorly understood. This paper presents a work-in-progress study that investigates these mechanisms through a multi-method design combining agent-based modelling (ABM) and fuzzy-set Qualitative Comparative Analysis (fsQCA). The ABM simulates how shared understanding emerges from the micro-level interactions of diverse agents under five conditions derived from a systematic literature review. The fsQCA then examines whether the configurational patterns generated by the model find empirical support across real-world cases drawn from the Collaborative Governance Case Database. The paper reflects on what this combination may offer as an analytical strategy for studying emergent phenomena in governance contexts.