Intelligence Analysis: Developing a Model of the Process Based on a Cognitive Task Analysis

Abstract

Intelligence analysts must filter, analyze, synthesize and determine the relevance of a continual stream of incoming information that often pertains to several different situations. The sheer volume makes it hard to process all the data, yet currently available technology is not always effective in helping the analyst assimilate the huge amount of information that needs to be synthesized. A critical part of the analyst's task also involves projecting future anticipated events. Products of this information foraging and analysis are used by senior decisionmakers to make high-stakes decisions. One goal for the research reported here is to capture data that will provide input to building computational models of the intelligence analyst and analytic processes used by intelligence analysts. A Cognitive Task Analysis (CTA) was conducted to support development of computational models of the analyst's processes, biases, and analytic strategies. CTA is an extension of traditional task analysis techniques to produce information regarding the know-ledge, thought processes, and goal structures that provide the foundation for task performance (Chipman, Schraagen, and Shalin, 2000). The goal of CTA is to discover the cognitive activities that are required for performing a task to identify opportunities to improve performance by providing improved support of these activities. (Potter, Roth, Woods, and Elm, 2000).