A personality test is considered reliable if it produces the same result repeatedly and valid if it really measures what it claims to measure, explains Kendra Cherry on About.com. A reliable psychological test is consistent and dependable.
Reliability means the consistency of a measure, states Cherry. For example, a test designed to measure a particular trait, like introversion, should produce approximately the same results each time it is administered to a subject. It is hard to calculate the exact reliability of a test, but it can be estimated in several ways.
To assess test-retest reliability, a test is usually administered twice at two different periods. This type of reliability determines the consistency of a test across time, and it is often used for stable traits, such as intelligence. Generally, a test yields higher reliability when only a short time has passed between the tests.
Another type of reliability is inter-rater reliability, which involves two or more independent judges who score the test. Comparing the scores determines the consistency of the estimates. Other forms of reliability are parallel-forms reliability, which is assessed by comparing two different tests that were made using the same content, and internal consistency reliability, which is used to evaluate the consistency of results of different items on the same test.