FORMS AND DIMENSIONS

In its broadest sense, collaboration in the IC occurs whenever one analyst seeks assistance from another analyst, outside contractor, or unpaid expert. Such collaborations vary along several dimensions that, when matched appropriately to situational needs, can improve their effectiveness. One is group duration, which can range from an ad hoc one-time arrangement to long-standing working groups. A second is the number of analysts involved. A third is how direct communications are, ranging from face-to-face meetings to anonymous electronic exchanges. Fourth, collaboration can be either cooperative (seeking shared conclusions) or competitive, maintaining alternative views either deliberately (e.g., red teaming or devil’s advocacy) or naturally (e.g., genuine disagreement).

Although all collaborations involve some integration of independent perspectives, consensus is not necessary for the outcome to be useful. Indeed, research in private corporations finds that people are most satisfied with collaboration that strives for “consensus with qualification,” in the sense that the final product reflects a majority view, while preserving dissenting views as “qualifications” (e.g., Eisenhardt et al., 1997). Useful integrated outcomes need not be produced by analysts themselves, but might be aggregated by an external team or even software “robots” (as in a prediction market). Box 5-1 describes an emerging form of parallel collaboration: “idea tournaments” for producing the best ideas for solving a problem.

Intelligence in the age of global counterterrorism requires effective collaboration with groups both inside and outside the IC, including domestic and international agencies, private contractors, industry experts, and academics. These relationships can range from informal calls for advice to formal contracts.

Assessment of the value of collaborations face the same evaluation challenges that arise with individually produced analyses (Chapter 2). Independent review committees can provide empirical evaluations (e.g., those conducted by the National Institutes of Health, National Science Foundation, Institute of Education Sciences, and Office for Judicial Research; for an example in private business, see Sharpe and Keelin, 1998). The Intelligence Advanced Research Projects Activity (IARPA) performs some of this function for the IC, by evaluating new high-risk/high-payoff methods with a small probability of producing an overwhelming intelligence advantage. However, IARPA does not support research on current analytic practices. The Office of Analytic Integrity and Standards in the Office of the Director of National Intelligence (ODNI), along with units in other IC agencies, intensely evaluates analytical work done internally, but typically their pro-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement