Search

This paper describes a scaled-world simulation developed to conduct empirical research on team cognition, communication, and decision-making within a distributed environment. The NeoCITIES simulation is an advancement of the CITIES task, which was designed to study group decision-making within a command, control, and communications (C3) setting (Wellens & Ergener, 1988). Studying group decision-making is a two-fold problem involving team cognition and team communication. According to McNeese (2003), team cognition is constructed through distributed and emerging activities via several sources. A majority of studies examining distributed decision-making have involved militaristic, battlefield engagement, or urban warfare settings. In that same spirit, NeoCITIES was designed for emergency crisis management teams undergoing terrorist attacks within a college-town. Thus, NeoCITIES is a new and operationally relevant scaled world that emulates the complexities and emergent decision-making attributes resident in a 9/11-type of terrorist scenario. Through the use of NeoCITIES, we anticipate the assessment of a number of cognitive tools to support distributed cognition (e.g., problem-based decomposition) as well as advancing adaptive intelligent interfaces.

This paper emphasizes the use of cognitive task analysis to gain significant insight into the unique domain of intelligence analysts, how intelligence analysts view this domain, and how this domain can be replicated in a controlled simulation environment in which innovative tools and procedures can be empirically tested. Details of two comprehensive knowledge elicitation sessions involving intelligence analysts are provided as an example of using the Advanced Knowledge Acquisition and Design (AKADAM) methodology to obtain contextually relevant information for use in developing a homeland defense-oriented simulation/experimental task. Several distinctive characteristics of intelligence analyst functionality were discovered, including the multi-source integration of relevant information, complex cognitive analysis, and team collaboration in decision-making. Additional themes such as social interaction and the limitations of current analysis tools were identified.

Complex problems such as analysis of military situation assessment, homeland defense, diagnosis of the health of complex systems, medical diagnosis, and environmental monitoring require the ability to utilize a wide variety of data such as signals, images, textual information, and scalar data. The rapid evolution of micro-scale sensors, wideband communications, and microprocessors enables the collection and dissemination of huge amounts of data to be provided to a human analyst. Unfortunately, the analyst cannot directly understand nor process the data. Instead, analysts reason about high-level abstractions via language. A challenge exists to decompose general problems into detailed models that link to specific types of data (viz., problem centered decomposition) and to compose data into meaningful relationships to assist the understanding of semantic representations of abstract concepts. This paper discusses the challenge of problem-centered analysis (including problem centered decomposition and problem centered composition) and describes our efforts to develop cognitive aids to assist the analysis process for improved understanding of complex problems.