Linear regression and the normality assumption

• Published in: Journal of clinical epidemiology Vol. 98; pp. 146 - 151
• Main Authors: Schmidt, Amand F., Finan, Chris
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.06.2018; Elsevier Limited
• Subjects: Bias; Big data; Computer simulation; Confidence intervals; Diabetes mellitus; Diabetes mellitus (non-insulin dependent); Economic models; Empirical analysis; Epidemiological methods; Estimates; Hemoglobin; Hypotheses; Impact tests; Linear Models; Linear regression; Mathematical functions; Modeling assumptions; Normal distribution; Normality; Regression analysis; Regression models; Researchers; Sample Size; Statistical analysis; Statistical inference; Transformations; Variables; Violations; Big data; Statistical inference; Epidemiological methods; Linear regression; Bias; Modeling assumptions

Researchers often perform arbitrary outcome transformations to fulfill the normality assumption of a linear regression model. This commentary explains and illustrates that in large data settings, such transformations are often unnecessary, and worse may bias model estimates. Linear regression assump...