Two sample unpaired T-test power calculation using simulation

Abstract

Statistical tests like t test are used to test the hypothesis for comparison of mean between two groups. The result from a t test is used to determine if there a significant difference between the two samples mean, which cannot be attributed to sampling error or to random occurrence. A t test is considered a parametric test, meaning test samples should meet assumptions of normality, equal variances and independence. The article aims to calculate the power of unpaired t test using simulation in Microsoft Excel.