Last’s Dictionary of Epidemiology says that a covariate is a variable that is possibly predictive of the outcome under study. Even that is somewhat wrong because the word predictive means that the covariate has to occur in time before the outcome that is predicted, but covariates are frequently used when the covariate follows the outcome, when it’s coincident with the outcome as well as when it proceeds the outcome.
To learn more from Dr. Helena Kraemer listen to the podcast episode below.
Click here for more Research Methodology Podcasts