Intelligence Quotient - Reliability and Validity

Reliability and Validity

Psychometricians generally regard IQ tests as having high statistical reliability. A high reliability implies that while test-takers can have varying scores on differing occasions when taking the same test and can vary in scores on different IQ tests taken at the same age, the scores generally agree. A test-taker's score on any one IQ test is surrounded by an error band that shows, to a specified degree of confidence, what the test-taker's true score is likely to be. For modern tests, the standard error of measurement is about three points, or in other words, the odds are about two out of three that a person's true IQ is in range from three points above to three points below the test IQ. Another description is there is a 95% chance the true IQ is in range from four to five points above to four to five points below the test IQ, depending on the test in question. Clinical psychologists generally regard them as having sufficient statistical validity for many clinical purposes.