Step 1 – Group discussion for filling in the cross impact matrix
Using figure 1, the impacts of each variable on the others are assessed during an individual brainstorming and the following group discussion. This can be done in different sub-groups to limit the time necessary for discussion.
For each variable, its influence on all other variables is considered: “If the variable X changes, how big is the impact on variable Y?”. The answer is given in the form of a score between 0 (no impact) and 3 (high impact).
Time necessary for ranking the relations is a useful indicator of the knowledge shared by the participants and the leader has to keep trace of this hidden (out of the matrix) information, in order to better address the following discussions (e.g. the interpretation of the system and the choice of the representative “active” variables..).
Figure 1: Cross Impact Matrix
Step 2 – Sum up the scores
In this step, the sums are calculated for the lines (active sum) and the columns (passive sum).
The individual behaviour of the variables is represented by the values of the active-sum (AS) and the passive-sum (PS). The active sum represents the ability of an individual variable to influence all other variables in the system, whereas passive sum is a corresponding value for the reaction of the variable due to changes of other variables in the system.
Step 3 – Visualisation and Interpretation of Cross Impact Analysis (CIA)
The results are depicted in a diagram that is segmented into different zones. According to the zone they appear in, the variables show typical characteristics.
Active variables allow effective changes in the system and thus have the potential to re-stabilize it in a new state. They are of major interest for the design process of PSS (Scenario Building, Worksheet 17).
Critical variables have to be handled with caution because they have big potential for driving and changing processes, but they can easily get out of control, or destabilize the system.
Variables in the reactive zone represent important indicators but have no steering potential.
Buffers have a limited effect on the system, and the neutral zone provides variables for self-regulation but again no good candidates for steering.
The variables represented by their line and column sums are depicted in the format figure 2:
Figure 2:Variable interpretation
Results are then interpreted via a group discussion to identify the most promising candidates among the variables that are useful for scenario building (Worksheet 17).