The Cognitive Sciences Laboratory was asked to contribute anomalous cognition data in an experimental mode to various segments within the U.S. Intelligence Community. Most actual operations were conducted by CSL prior to 1978 when the government set up an in-house unit for this purpose; however, some few operational AC trials were conducted through 1994.
Since this activity was on an experimental basis, the CSL was also asked to participate in "test-bed trials" to assess the validity of the technique in a variety of circumstances. (See examples at bottom of page.)
Operational anomalous cognition differs considerably from laboratory trials in which strict protocols allow for valid statistical analyses. In the real-world, the goal of operations is to obtain new and useful information. The intelligence value of an AC trial, however, is only loosely connected to the quality of the AC data. For example, a near perfect AC description of a foreign site might be of little intelligence value because all the information was known from multiple other sources. On the other hand, an AC trial that would otherwise not score well in the laboratory, might provide clues to an analyst that actually cracks the intelligence problem.
The protocol for operations also differ from those in the laboratory. It is a research issue to determine how much a-priori information should be given to the viewer to define the range of possible targets and when it should be given. In the laboratory, we found that some information is optimal; whereas too little and too much may lead to incorrect results. The details can be found in our published account of the research that led to this conclusion, "Managing the Target-Pool Bandwidth: Possible Noise Reduction for Anomalous Cognition Experiments." The degree to which this result transfers to the operational setting remains to be determined.
We have already mentioned that operational anomalous cognition need not be very accurate to be useful. It is a general rule for the use of intelligence data: neither policy nor action should be taken as a result of a single source of intelligence. CIA Analyst "Norm" said on ABC Nightline, November 28, 1995:
He was correct, and the implication is that AC-derived data was useful to the intelligence community.
We present a few examples of anomalous cognition trials that addressed questions of National Security.