In this paper we present a statistical method for the identification of causal factors in a complex process. The method is based on the use of the correlation coefficient and of the correlation matrix rank, in the goal to determine linear relations between huge volume data sets. The above mentioned relations can drive to the verification of some hypothesis or to determine new connections or links between different phenomenon. The paper also presents an example for the method. The interested domains, from this method point of view, are diverse and important: military, social, medical, economic, etc.