Compare Means: Parametric Test vs Non-parametric Test: Independent sample t-test

How do you treat outliers in your datasets?
You need to understand the reason for your outliers before treating or removing outliers in the datasets. The presence of outliers in the datasets is one of the reasons why your datasets are not normally distributed. Parametric test is for a normally distributed datasets while, non-parametric test is for a dataset that are not normally distributed.
So Instead of removing an important outlier, why can't you opt for a non-parametric test?.
This video explain compare means test, which is usually used to compare the means of two or more groups and also to check if their differences are statistically significant.
Compare means techniques is usually for a parametric test while, compare median techniques is for non-parametric test. This video also explains various compare means estimation techniques such as independent t-test, one sample t-test, paired t-test and one way ANOVA test.