ESSA Prize assigned: Jessica Gu, Cristina Cecchini.

Best student poster:

Cristina Cecchini (UNIVERSITY OF FLORENCE) for Deliberate Self-harm: a study about the evolution of stable maladaptive strategies. (Cecchini Cristina, Meringolo Patrizia, Guazzini Andrea).

Self-injury is a maladaptive behavior described as the intentional injuring of one’s own body without suicidal intent, and it is very common in adolescents. The Experiential Avoidance model claims it is a negatively reinforced strategy for terminating unwanted emotional arousal, which reveals complex dynamics. Literature has not revealed the role of diverse factors in affecting such a behavior. The aim of this study is to model dynamics of self-injury in adolescence, through an agent based modelling (ABM) approach, by focusing on network topologies (i.e., Uniform, Gaussian, Exponential), three main categories of risk factors (i.e., Inner Factors, Outer Factors, Media Factors), and the interaction between nodes. A probability to experience stressful events is fixed, and the final number of self-injurious agents is the order parameter. Results are expected to show the combine effect of risk factors and topology, highlighting interesting scenarios about the complex dynamics of the phenomenon.

Best student paper:

Jessica Gu (THE UNIVERSITY OF TOKYO) for Agent-Based Simulation of Organizational Knowledge Management (Gu Jessica, Chen Yu)

An agent-based model is employed to simulate an organization as a complex adaptive system which reveals how it creates value through evolutionary knowledge management by autonomous agents from the bottom up. In the design, agents solve problems and improve performance under an uncertain environment through two key knowledge management (KM) strategies: creating new knowledge independently – Innovation or acquiring shared knowledge through creating social networks – Imitation. Agents preserve free-will on choosing either KM strategy for problem solving and selecting other agent to learn from. The endogenous likelihood of choices is updated with experience-weighted-attraction learning rule. The productivity of new knowledge and the connectivity of the social network are two exogenous factors that are controlled by the policy maker for administrative intervention. In the simulation, various causal relations among agents’ behavior, the turbulence of environment, the emergent social structure and the organizational performance are elucidated. One of the surprising findings indicates that long-term steady-state organizational performance is non-monotonically improved by increasing social network connectivity under turbulent environment while monotonically improved under stable environment. To verify and improve the developed agent-based model, a series of behavioral experiments are conducted with human participants. After comparing results obtained from both simulation and experiments, a scarcity heuristic has been identified suggesting a mental shortcut and cognitive bias of human agents who place a value on low probability event. Based on such findings, the modification of the agent-based model has been proposed and tested. Taking agent heuristic decision making into consideration, the agent-based model can be more sophisticated and advanced, hence, better KM policy can be made for organizational performance optimization in the future work.

← Older Newer →