Empirical Inference of Numerical Information into Causal Strategy Models by Means of Artificial Intelligence - Distributed Artificial Intelligence, Agent Technology, and Collaborative Applications - page 307

Figure 6. Generating causal system

comprises the nomologic cause-and-effect hypotheses as assumed a priori by decision makers. For this
purpose the functional proof of concept starts with a fully interlinked causal system consisting of the
five variables already mentioned.
In order to rule out spurious associations according to the second and third causality criterion (infor-
mational redundancy and temporal precedence) the proposed approach analyzes the cross correlations
for every link in the maximal strategic system model for different time lags. Since fourteen associations
do not show a significant cross correlation at any ∆ t , they are marked as spurious. The remaining six
potentially causal relations according to the second and third causality criterion yield cross correlation
patterns as shown in Figure 7.
As it is obvious from this correlogram, the proposed proof of causality identifies the correct win-
dows of impact of the five causal relations implicitly included in the generating system. However, this
algorithm untruly identifies a sixth association ( SP → CG ) to be potentially causal according to the
second and third causality criterion because its cross correlation at a time lag of seven periods seems
to be significant:

.

(7) 0.1458 = > = 0.1411

SP CG , (7)

SP CG ,