The three principles of test construction include standardization, reliability, and validity. In order for a test to be useful, it must contain all three elements. Standardization refers to the meaningfulness of test scores among test takers. It also includes providing all test takers with the same instructions, questions, and time constraints. A standardized score tends to follow a bell curve of score distributions and determines where a test taker's performance is relative to other testers. Secondly, a test must also be reliable to be useful. Reliability means that the test produces consistent results. A reliable test will give the same or similar score when someone is tested with an alternate version or retested with the same test. This establishes test-retest reliability. In order to establish split-half reliability for a test, even and odd problems may be designed to measure similar aspects and are then scored separately. This allows us to see if the person is consistent in scores on each half of the test. Finally, a test must be valid and measure what it is designed to measure. It is possible to have a very reliable test that is not valid.
Responses may vary but should include some or all of the following information:
Reputable tests have three factors in common. They are all standardized, valid, and reliable. Standardized tests have scores that are compared to the results of other test takers. Valid tests cover the information that is being looked at. Reliable tests have results that remain consistent over multiple testings.
