A Modulo-Based Computation for Kendall’s Phi Test for 2 by 2 Contingency Table

Kendall’s Phi Test is very important technique in testing whether there exists a significant relation between variables under some known levels of experiments. However, the typical statistical testing would utilise some complex probability distribution for the statistic. In this article, we put forward a computational approach to effectively linearize Kendall’s Phi coefficients based on modulo operation. In our approach, the confidence levels would serve the threshold of the operation. Unlike typical approach which lists the whole distribution of the statistic, our method would concentrate only on the accumulated values needed. Such computation has merits in visualization and programming. In particular, if one is interested in Kendall’s Phi-based test, then this approach could largely reduce the complexity and computational time.