The INSIDE assumption under all positive coding: interpretation and partial empirical assessment
Mendelian randomisation (MR) implemented through instrumental variable (IV) analysis is a popular strategy for strengthening causal inference in observational studies. A key assumption for several MR estimators, including MR-Egger regression, is the INstrument Strength Independent of Direct Effect (INSIDE) assumption....