This occurs when there is systematic error in the techniques used to measure or record data – this could be for exposures/risk factors, outcome of interest (usually disease state), or extraneous variables (potential confounders).