AN EXTRACTION OF KEY VARIABLESFOR LONG-TERM TRENDS USING A MODIFIEDCROSS-IMPACT MATRIX
1232-9312
Abstract
The traditional approach to forecasting lies in the assumption that, byapplying a set of powerful analytic tools, the future could be predictedaccurately enough to choose an appropriate strategic direction for it. Theprocess often involves underestimating either expert knowledge or quantitativemethods side in order to lay out a vision of long-term future to be captured inscenarios. Computer science offers a variety of methods and computer tools tofacilitate the process of strategy formulation.The identification of impactful changes, called “key variables,” “keyfactors,” “rely variables” or “driving forces,” within the environment that arelikely to take place helps to reduce the uncertainty of the future and could beutilised throughout the strategy making process. Key variables shaping the longtermfuture are identified by some of the following questions: What are thedriving forces? What is uncertain? What is inevitable?This paper demonstrates an original approach to variable selection bymodelling a comprehensive dataset of variables using a wide variety of bothtechnical and non-technical parameters indicated by experts. The extraction ofkey variables that significantly influence future changes of technological,environmental, political, and social phenomena is done with the modified crossimpactmatrix. The proposed method contains a sequence of qualitative andquantitative procedures to overcome some obstacles that occur within theclassical cross-impact method.