Alternative Methods of Solving Biasedness in Chi – Square Contingency Table

Abstract

Some methods of solving the biasedness in Chi–square Contingency table statistic were considered. Phi Coefficient, Contingency Coefficient and Cramer’s V tools were employed to solve the biasedness in the use of Chi–square test. Our results show that any of Phi coefficient, Contingency coefficient or Cramer’s V can be used to describe the association between two variables if the data matrix is 2 x 2. Contingency Coefficient was recommended as a good statistic when the matrix dimension is the same while the Cramer’s V is most adequate when the data matrix differs.