Structured Analysis in intelligence, not to be confused with the software engineering practice by the same name, refers to structured methods used to model and normalize the intelligence analysis process. The structured approach to analysis has several main benefits:

1) Empirical Foundations: Structured Analysis focuses on empirical (evidence based) foundations vice relying on opinion or “expertise.” These methods fight the cognitive biases and preexisting beliefs that plague the human thought process effectively tainting our view of the world and our assessment of it. The most commonly utilized architecture centers around hypothesis testing models. The most widely known and utilized structured method is Richard Heuer’s “Analysis of Competing Hypotheses.” This method forces the analyst to weigh each piece of evidence indicating whether it is consistent or inconsistent with the different competing hypotheses. It also emphasizes the search for evidence that supports a competing hypothesis verses searching only for information that support a primary hypothesis. This method effectively combats cognitive biases such as Confirmation Bias, Anchoring Heuristic, Cognitive Dissonance, etc.

2) Audit Trails: Structured Methods provide audit trails for analysis. Structured methods provide detailed information about how an analyst came to his/her conclusions and assessments; how he/she measured and weighed each piece of evidence; and how the entire body of evidence adds up to indicate a particular situation or future event. This allows other analysts and supervisors to review analysis; reconsider the evidence in different contexts or as more evidence comes to light; and/or to look back at assessments to determine what went right and what went wrong. For automated systems the audit trail is particularly important. Automated Intelligence Analytics can collect, organize, analyze, weigh information to produce intelligence analysis; however, it is important to be able to display how or why the automated system has come to its conclusions so human analysts and decision-makers can review the intelligence, confirm the results and develop strategies to act on the intelligence. No one wants to trust and act on what a computer tells them to do; Structure Analysis allows the user to be confident on its findings because he/she can clearly see how and why a body of evidence leads to a particular conclusion.

3) Comparable Analysis: Structured Analysis produces comparable results across multiple intelligence efforts. Structured Analysis techniques use the same methodology and evidence weighting for each target of intelligence and each intelligence effort utilizing the structured method. Therefore, the results of one intelligence effort are comparable to the results of other efforts utilizing the same structured technique. This not only normalizes results across multiple efforts, but also normalizes assessments of multiple analysts. These practices increase the confidence of decision-makers in assessments of their intelligence assets and allow them to understand how one result compares to another. For instance, if one assessment asserts that event A is highly likely to happen in a particular time frame and another assessment asserts that event B is highly likely to happen in a separate timeframe, the decision-maker can be confident that the evidence in both analyses have the same strength of indication. This is particularly important in large datasets with multiple intelligence targets.

4) Scaled Indication (likelihood): Structured Analysis uses standardized scales of likelihood. One example of this is the “Words of Estimative Probability Scale” or WEPS. WEPS, adopted by the National Intelligence Council (NIC), utilizes seven descriptors that identify seven levels of likelihood: “All But Impossible;” “Highly Unlikely;” “Unlikely;” “Even Odds;” “Likely;” “Highly Likely;” and “All But Certain.” WEPS and other scales of indication/likelihood make sure all analysts are speaking the same language and ensures the definitions used to describe the level of likelihood represented by a body of evidence is consistent among all analysts. Scaled Indication is necessary for automated analytic systems in order to dovetail with human analyst’s Standard Operating Procedures (SOP). Automated systems need to be able to produce intelligence that is scalable to human analysis and produce results easily understood by human consumers. For example, if we assume that any two WEPS have an equal increment of likelihood between them, then each WEP represents a .1666 or 1/6 (six increments between the 7 WEPs) increase in likelihood over the previous WEP. Analytic algorithms can develop an indicative or estimative value of a body of evidence that can seamless dovetail with human analysis. DAGIR Co. Analytics use a variety of Indication Scales that vary from scales as simple as a 1-10 Red Flag Scale to complex scales such as statistical probability scales (i.e. p = .0286). The scale used in a particular DAGIR Co. Analytic Systems is generally determined by analyzing your organization’s SOP, common language, and/or industry standards.

5) Analytic Modeling: Structured Analysis allows DAGIR Co. analysts to model your organizations dilemma-space surrounding its areas of uncertainty. Organizations have continuous or reoccurring situations and environments where their decision-makers must devise strategies and guide operations under conditions of uncertainty. These situations and environments have indicators or regularly expressed evidence that indicate events, circumstances, behaviors, and/or other eventualities. These indicators of eventualities or potential outcomes maybe precursors, functional requirements, relationships of situational/environmental or temporal elements, and/or correlates of said outcomes. DAGIR Co. Structured Analysis techniques allow its analysts to model these areas of uncertainty and build data mining apparatuses to detect, collect, and organize all indicators that present themselves. Then, using analytic algorithms to measure, weigh, and analyze those indicators, DAGIR Co. Analytics develop full-scale analysis and will keep doing it indefinitely 24 hours a day.