[1] The Central Intelligence Agency (CIA) became interested in the methodology in the late 1960s and early 1970s as an analytic technique for predicting how different factors and variables would impact future decisions.
[4] The basic principles of cross-impact analysis date back to the late 1960s, but the original processes were relatively simple and were based on a game design.
Theodore J. Gordon writes that cross-impact analysis was the result of a question: "can forecasting be based on perceptions about how future events may interact?
[5] In 1974, Duperrin and Godet developed Cross Impact Systems and Matrices (or SMIC) in France for prospective forecasting studies.
In 1980, Selwyn Enzer at the University of California incorporated cross-impact analysis into a simulation method known as Interax, The Delphi technique was combined with Cross Impact Analysis in 1984, and researchers at Texas A&M University used Cross Impact in a process called "EZ-IMPACT" that was based on Kane's algorithm from KSIM.
Cross-impact analysis was being used to solve real world issues as John Stover applied the methodology to simulate the economy of Uruguay.
These relationships are then categorized as positive or negative relative to each other, and are used to determine which events or scenarios are most probable or likely to occur within a given time frame.
[4] Researchers must calculate the numerical values or percentages very precisely to ensure accurate results and that impacts of events on each other are realistic and not contradictory.
The conformity of the style generates a certain level of inflexibility when dealing with variables other than events, like environmental conditions or political issues.
Shortly after Theodore Gordon and Olaf Helmer developed the original cross-impact method, the United States intelligence community picked up the technique and has been using it for over thirty years.
[2] While the basic premise of relationships and impacts between multiple variables remains the same, the intelligence community modified cross-impact analysis to meet its various needs.
[15] In addition, intelligence analysts can choose to use more flexible measurements like "enhancing", "inhibiting", or "unrelated" instead of the rigid mathematics of the tradition methodology to include non-event variables.
[17] While several traditional cross-impact analysis methods suggest the creation of a matrix, the priority still relies in probabilities, one-to-one relationships, and the order of events.
This also enables experts in a topic to use the methodology relatively quickly without having to cross-check the numerous calculations faced by the Futurist Forecasting Style.
The flexibility of the style relies heavily on the opinions and knowledge of the analysts involved, and is difficult to reproduce results with a different group.