Cognitive biases and debiasing in intelligence analysis

Herbert Simon's ideas on human rationality have framed a generation of psychological research into decision making. Researchers have identified numerous "cognitive biases"––departures from normatively rational behaviour––and they have begun to develop and test debiasing interventions to combat those biases. Simon believed that individuals working in an organizational context could be helped to achieve rational thinking by the organization itself. This chapter considers how organizations responsible for intelligence analysis help their analysts to overcome cognitive bias. Intelligence analysis is a cognitively challenging task that is performed under suboptimal conditions, and analysts are often portrayed as being uncritical and suffering from cognitive bias. Eight biases are identified that may manifest at various stages of the analytic workflow (i.e., belief bias, confirmation bias, explanation bias, fluency effects, framing effects, order effects, the planning fallacy, and overconfidence). The intelligence community's response to the potential problem of cognitive bias has largely been to adopt ad hoc, untested, "structured analytic techniques" and computer technologies. Intelligence organizations have largely eschewed psychologically informed and empirically tested debiasing interventions. It is argued that, as such, these organizations may fail to fulfil their function in terms of providing analysts with an environment that encourages rational thinking.