The extent to which the design, conduct and analysis of a study are likely to have prevented systematic error and to have produced results that are reliable and valid.