The article compares different tests used to check if the average values of related data sets are the same. When the data is in the form of Likert scales or not normally distributed, non-parametric tests are used. The study found that for small sample sizes, parametric tests like t -test and F -test are better at controlling errors and have more power. But for medium and large sample sizes, both parametric and non-parametric tests perform similarly. The power of both methods increases with larger sample sizes and stronger correlations.