I'm a current undergraduate student. I often get asked source when I explain that linear regression need only residual normality assumption, and not variable normality. I know there's so many website that explain this on the internet but they don't believe this kind of citation. Any recommendation about maybe journal artical or something similar that I can refer them to?