In general, a systematic (i.e., not due to random error) deviation in an observation from the true nature of an event.