A systematic error in measurement that causes differential performance in certain groups on individual items or the overall test, such that inferences regarding the trait of interest are adversely affected. Affects validity.