It is a common practice in data analysis to remove features (independent variables) with low variance for dimensionality reduction, with the justification that a feature with lo