Hands-on Tutorial
The Simulation of Bootstrapping for Confidence Interval and Hypothesis Testing
How to generate bootstrapped samples for calculating the confidence interval and two-sample hypothesis testing
We have already known that the most common statistical technique used for hypothesis testing is known as parametric statistics. It is based on the distribution assumptions like Normal distribution or other distributions. What if our data doesn't follow any distributions for the basic principle of using parametric statistics? Can we perform other statistical techniques? Yes, of course! There are still a lot of techniques you can easily use, one of them is bootstrapping.
This article will cover what, why and how we should use bootstrapping for data analysis. You can also try the bootstrapping simulation on your own computer using the Jupyter Notebook. It is enclosed in the last section.
Without further ado, let’s jump in!
The importance of bootstrapping
Bootstrapping is a random sampling technique with a replacement in the objective to estimate the parameter of a population. Suppose that we have a large population with distribution D. It has elements X1, X2…
