Intelligence Application Tools
While all intelligence topics are unique and require both current information and an up-to-date context, some issues recur over time and are relevant in a variety of different situations. Over the years EBR has created tools that provide frameworks and analytic structures that support intelligence analysts when faced with these recurring intelligence issues.
Rather than replacing the judgment of country analysts and regional specialists with statistics, the EBR approach is to embrace their knowledge as the most current and broadly informed sources of information and knowledge. Hence, these applications are expert driven models that use informed responses to inform structured analyses. These expert driven models are often used not merely to make projections, but also to allow structured discussions and debates among the experts that illuminate the key factors involved in a particular application. Many of them generate contingent forecasts that use the rich understanding of the context and situational dynamics to clarify possible alternative futures and the circumstances that make them more or less likely.
The life cycle of these projects reflects the best practice of applied science. The first step is a systematic review of the existing literature and determination of the relevant approaches or schools of thought and how they are similar and different. The second step is a set of open ended interviews with recognized experts from across the range of potential applications. This phase allows recognition of differences (e.g., cultural factors) and underlying similarities. Based on these efforts, a set of hypothesized causal relationships is created, representing alternative models. These hypothesized sets of relevant variables and relationships are then tested empirically using inputs from a variety of country and regional experts. In some cases historical “post-diction” is also used to test these hypotheses, but that approach is less valid than current applications in which the experts do not know the outcomes of the cases of interest. This phase can be lengthy, one year or more of effort is often needed to determine which of the hypothesized patterns are empirically valid and which variables can be successfully estimated by the experts. The results of these efforts are then subjected to statistical tests to determine the most efficient and cost-effective patterns for incorporation into the model. This model is then subjected to further testing and is used as the basis for in-depth interviews with country and regional experts. Only after this validation phase is completed does EBR certify the final tool. Along with the software embodiment of the tool, EBR also creates a training course and support package that can be used by those adopting the tool and approach.
While this approach requires time, effort, and cooperation from the community of users, the resulting models have proven cost effective and robust across a wide variety of geographic, cultural and political contexts.