Author Biography

Dr. Gordon R. Middleton is a retired Colonel in the U.S. Air Force, with over four decades of management, intelligence, and academic experience, including executive experience in the NRO, USAF, ODNI, FBI, CIA, and NSA. He is a DOD-certified Program Manager (Level III), who served as program manager for an Intelligence Community Category IA major program. Currently, he is Director of the Strategic Intelligence Program at Patrick Henry College, and a commissioner on the TRACS Accreditation Board. He previously served as Director of Analysis at Business and Engineering Systems Corporation, 2007-2014.



Subject Area Keywords

Economics, Intelligence analysis, Intelligence studies/education, Methodology


In an era of rapidly increasing technical capability, the intelligence focus is often on the modes of collection and tools of analysis rather than the analyst themselves. Data are proliferating and so are tools to help analysts deal with the flood of data and the increasingly demanding timeline for intelligence production, but the role of the analyst in such a data-driven environment needs to be understood in order to support key management decisions (e.g., training and investment priorities).

This paper describes a model of the analytic process, and analyzes the roles played by humans and machine tools in each process element. It concludes that human analytic functions are as critical in the intelligence process as they have ever been, and perhaps even more so due to the advance of technology in the intelligence business.

Human functions performed by analysts are critical in nearly every step in the process, particularly at the front end of the analytic process, in defining and refining the problem statement, and at the end of the process, in generating knowledge, presenting the story in understandable terms, tailoring the presentation of the results of the analysis to various audiences, as well as in determining when to initiate iterative loops in the process.

The paper concludes with observations on the necessity of enabling expert analysts, tools to deal with big data, developing analysts with advanced analytic methods as well as with techniques for optimal use of advanced tools, and suggestions for further quantitative research.