Developing Metacognitive Models for Team-Based Dynamic Environment Using Fuzzy Cognitive Mapping Jung Hyup Kim, Gretchen A. Macht, Ling Rothrock, and David A. Nembhard Department of Industrial and Manufacturing Engineering, Pennsylvania State University, USA {jzk170,gam201,lrothroc,dan12}@psu.edu Abstract. In this paper, by using Fuzzy Cognitive Mapping (FCM) technique, we developed the metacognitive models for team-based dynamic environment. Preliminary findings from our metacognitive studies provided a possible meta- cognitive framework in dynamic control tasks [1, 2]. By analyzing metacogni- tion, performance, and communication data between team, we are able to develop the team-based evolving metacognitive models for the dynamic envi- ronments using a fuzzy cognitive map. In this research, a human-in-the-loop simulation experiment was conducted to collect communication data, objective performance data (operator on-time action performance), and subjective rating data (retrospective confident metacognitive judgment) from 6 dyads (12 partici- pants). Within the Anti-Air Warfare Coordinator (AAWC) simulation domain, the simulation test bed provides an interactive simulating condition in which the monitoring team must communicate with their team member to defend their ship against hostile aircraft. Keywords: Metacognition, Team Performance, Human-in-the-loop simulation, Fuzzy Cognitive Map. 1 Introduction The need to develop more advanced on-the-job training methods has become a grow- ing concern in many industries because people not only work as individuals, but as members of teams. Although advanced technology provides the ability to develop more effective training approaches to novice workers, building effective team training methods is still on-going research area. To address this critical need, this research investigated the different behavior models based on metacognition and communica- tion between team members using Fuzzy Cognitive Mapping (FCM) techniques. FCM is a “mental landscape” of the elements (e.g. actors, values, goals, and trends) in a fuzzy feedback system. FCM can demonstrate the links between causal events of dynamic tasks and human behavior with the change of time. The map lists the fuzzy rules related with events to show causal flow paths like Hasse diagram [3]. The FCM enables job trainers to evaluate trainees’ internal learning states, and to help the in- structors to choose several possible actions. In this research, we developed a human- in-the-loop simulation, which emulates a computer-based dynamic control task to D. Harris (Ed.): EPCE/HCII 2013, Part I, LNAI 8019, pp. 325–334, 2013. © Springer-Verlag Berlin Heidelberg 2013