Common-Method Variance

Common-method variance is a measurement problem in applied statistics and psychometrics in which the correlation observed between two variables is inflated or deflated because both were measured using the same method or gathered from the same source, rather than reflecting a true relationship between the underlying constructs the measures are meant to capture. A simple illustration is an electronic survey interface that affects the responses of people unfamiliar with taking surveys online differently than it affects experienced users, introducing a shared source of error into both measures collected through it. Whether two particular variables actually show a meaningful common-method effect is not automatic and depends on both the specific measurement method used and the nature of the constructs being measured, so the bias does not uniformly distort every variable gathered through a shared instrument.