Exerpt from Clauser, J. K. and Weir, S. M. (1976).
Intelligence Research Methodology
Defense Intelligence College.
PREDICTION, FORECASTING, AND HARUSPICY
With the intelligence community's overriding concern for reducing uncertainty, it is not
surprising that techniques for anticipating future events or conditions would have a special
attraction for the intelligence researcher or analyst.
This chapter discusses selected techniques
for anticipating future events or conditions that have been (or can be) used in intelligence
research and analysis.
The first part of this chapter discusses the nature of prediction and
forecasting, discusses assumptions underlying prediction, and describes the relationship of a
technique to the nature of the phenomenon being considered.
Although this chapter is concerned with methodologies for forecasting, there is nothing
sacrosanct about the methodology which limits its application to predicting or anticipating future
Simulations, for example, can be used for diagnostic purposes as well as for predictive
That accurate predictions have been made on the basis of "non-predictive" techniques,
as such, is not only possible, but is also highly probable.
Types of Phenomena and their
Relationship to Forecasting
At the outset, it should be apparent that some types of phenomena are easier to forecast
At one end of the continuum, for example, are short-term,
occur at specific points in time.
These are the most difficult
types of phenomena to predict because they may not have any precedents.
unfortunately, these are the types of events that are of most concern to the intelligence
An example of this type of phenomenon would be an unexpected coup that toppled
a government overnight, or the defection of a key scientist on a highly classified defense-related
project, or a discovery of large deposits of a strategic commodity.
In the middle of the continuum are events which occur over a period of time.
There may be
little doubt about their outcome, but the concern may be to establish when the outcome will
occur or what the effects may be.
Admittedly, the point in time at which a phenomenon started
may be considered a unique event, but the event is nevertheless continuous for a given period of
An example of a continuous event would be a country's construction of its first aircraft
The precise time that the carrier was started (e.g., the specific point in time when the
first line was drawn on a plan) was a unique event.
But, for the sake of this discussion, the time
events and occurrences that would take place until the carrier completed its sea trials and became
fully operational could be considered as a continuous sequence of events.