5 Myths about Income Inequality Debunked | Dustin Carnes 
Statistical analysis of this subject might, at first glance, lead to exactly the conclusion that
discrimination is the primary culprit in income disparities between minorities and gender.
There are two problems with this.
First, it relies on faulty statistical interpretation. When scholars do a regression analysis—a statistical method where outputs like income can be predicted by independent variables—there is often a residual disparity leftover. This leftover unexplained disparity is taken as proof of discrimination. In reality, this unexplained disparity is the upper limit of all other unaccounted-for variables plus discrimination, if any. To automatically assume the remaining disparity is discrimination is a miscarriage of statistical interpretations of regression analyses.
Second, when we talk about a theory in science, we may have several competing theories trying to explain the same phenomenon. The best theories are often the ones that explain the most-observed phenomenon accurately, that make predictions, and that can be tested and verified repeatedly. The theory on/about discrimination in income disparities suffers from not being able to explain as many observed phenomena. It cannot make accurate predictions, and most importantly, it cannot be measured, tested, and verified. It is thus a theory that needs to be discarded when a superior one exists. . .
Go to https://fee.org/articles/5-myths-about-income-inequality-debunked/  to read the first four myths.
Myths originate by concluding from incomplete knowledge.
Myths disappear when theories are tested and verified.