Reliability refers to the consistency of test scores. It assures language testers that steady, accurate results are being achieved from one repeated measurement to another. For example, if we intend to measure the reading ability of a group of students, the test results we obtain should be approximately the same for the group over short periods of time and over parallel forms of the test. Comparable to validity, reliability is a key ingredient in assessment since it provides dependability and confidence in testing procedures-a component that is pivotal in the entire process of evaluation. Unlike validity, however, reliability is statistical in nature. This means that it can be calculated by statistical procedures.