A second‐class science? A defence of observational epidemiology to make causal inferences. (January 2014)