Reliability is the extent to which a test is repeatable and yields consistent results. Further more, all measurement procedures have the potential for error, so the aim is to minimize this error. Reliability can be improved by getting repeated measurements using the same test and, getting many different measures using slightly different techniques and methods. That being said an observed test score is made up of the true score plus measurement error and therefore is the most highly reliable. Simply put, teacher observation of a student is the most reliable measure.