Essentials about statistical bias
Math&Stats | Zero Bulls*it #3
The irrationality of a thing is no argument against its existence, rather a condition of it.
- F. Nietzsche
This article is a brief introduction into different types of statistical biases, its impact on the data analysis and decision making and how we can be aware of all this irrationality out there.
Introduction
The main purpose of this article is to show the nature of various statistical biases, its importance in statistics, data analysis, research and machine learning (ML) and give some idea of how to avoid it. There will be some examples of each bias type not only related to data analytics and ML, but also from the everyday life.
Read till the end to understand, why biased statistics are bad statistics and “Why cats who fell from higher floors have fewer injuries than cats who fell from lower down?”.
Intro
To get started, let’s get the idea of what a statistical bias is and why is it a crucial component of every data mining, data analysis and machine learning process.
What the hell is “Bias”?
To answer this question we will talk about bias from two perspectives: cognitive and statistical.
