Difference between observed and true scores due to random errors introduced by the measuring instrument or administering person, i.e., that portion of a score variance that can not be accounted for systematically (Note: Do not use for test bias errors)