Overview

Under the assumptions of one sample t-test, due to the definition of $p$ value, the False Positive Rate (FPR) should be 5%. This program will let you violate the normality assumption and see the resulting FPR. It will take $200,000$ independent samples from a chosen distribution with a chosen sample size and then perform one sample t test on them with the $h_{0}$ being true, meaning the value to test the difference from is the true mean of the distribution. It will then divide the number of significant $p$ values by $200,000$. This last quantity is, by definition, the FPR.

Purpose

This program has no practical benefit whatsoever 😝. It is merely a mean for the author to learn.

Link to the app

Click HERE to view the app