THE BEST WAY TO CHOOSING THE MULTIPLE COMPARISON TESTING FOR EQUAL VARIANCE AND UNEQUAL SAMPLE SIZE IN ONE WAY ANOVA

Abstract

Background: One-way ANOVA is a method for comparisons of three or more groups of continuous data. Multiple comparison analysis (MCA) is to identify significant differences between subgroups of study. In this paper were considered the comparison and choose the best multiple comparison testing for equal variance and unequal sample size in one-way ANOVA.

Materials and Methods: The most commonly used multiple comparison analysis is the Tukey, Scheffee, Bonferroni and Dunnett T analysis method. From the results, Tukey HSD, Scheffe, Bonferroni and Dunnet T procedure, showed a significant difference in BMI between definite & normotensive of blood pressure and between definite & borderline of blood pressure.

Result: The width 95% Confidence Interval of Scheffee procedure is higher than Bonferroni, Tukey HSD, and Dunnett T procedure

Conclusion: MCA tests may be necessary and researchers should think carefully about the many tests that should be used. This is because different tests can lead to different conclusions and careful consideration for appropriate testing should be given in each case.

Keywords: One-way ANOVA, Multiple comparison analysis (MCA), Equal variance, Unequal sample size and Width of 95% Confidence Interval.