A Correlation Model to Analyze Dependent Variables.

Abstract

In this paper a methodology is formulated to study the dependency between two variables. Statistical correlation coefficient between two variables is used as a measure of the degree of dependency. The method of bootstrap technique is employed to account for the statistical uncertainty in the correlation coefficient that is estimated from available data. A computer program entitled correlation coefficient generafor (CCG) is developed to perform the analysis. An example is also presented to demonstrate the methodology. The objective of this example is to determine the dependency between specific actions of control room operator of a nuclear power plant. The methodology and the CCG code are very effective and easy to use.