What is a standardized regression coefficient?

A standardized regression coefficient removes the original unit of measurement for variables in a regression equation. These coefficients are standardized and converted to a scale from 0 to 1. Since the values are standardized, a researcher can more easily compare the effect sizes of variables measured on different scales.

Standardized regression coefficients are also referred to as beta weights. A beta weight refers to the number of standard deviations an outcome variable will increase due to one standard deviation increase of a predictor variable. Critics of standardized regression coefficients argue that removing the unit of measurement from a variable can create bias due to sampling error.