This study demonstrated and compared the use of t-distribution, z-transformation and the Statistical Package for Social Sciences (SPSS) methods in testing the significance of the correlation coefficients. The study utilized data extracted from Eze (2019). The findings revealed that the significance of the correlation coefficient by the Fisher z-transformation and t-distribution methods were not at variance with that of SPSS. Results also showed that the SPSS approach appears to be the best method since it saves time, reveals the correlation coefficient and p-value for decision making. Results further showed that the t-distribution can be used for both large and small sample sizes since it approximates the z-distribution as the data tends to infinity. Based on the findings of the study, it was recommended among others that the SPSS approach should be adopted by researchers and data analysts for testing the significance of correlation coefficients because it is not only robust, but it is easier, saves time and energy; and users of SPSS in computing correlation should not go further to convert the derived correlation coefficient using either the t-distribution or the z-transformation to test a hypothesis to minimize the risk of computational errors, wrong decisions, time wastage and to avoid unnecessary repetition
Keywords: Correlation, Pearson, Relationship, Significance, Spearman
