Psychological testing is like any powerful tool -- in the hands of an honest, trained professional it can provide great benefit. It also has the potential to do a lot of harm. When I have taught classes in psychological testing I generally start out by asking students if they have any negative feelings about testing. Almost invariably one of the students will relate that a counselor in high school told them that their test score precluded success in college. And of course, the fact that this student is in my class, a graduate student and a college graduate, is testimony to the counselor's mistake.
I point out to the class that the test is not at fault, but that counselors and other test users frequently misuse the tests. Harm is done when tests are either taken too seriously or are disregarded. Understanding statistical concepts such as reliability, validity, correlation, standard error of measurement, and standard error of estimate is necessary for the proper interpretation of test scores. Generally licensed psychologists are the only professionals who have sufficient training to correctly interpret tests.
Take, for instance, an IQ score. IQ scores are pretty good predictors of academic ability. If an individual obtains an IQ score of less than 80, we could guess that he would not be successful in college. But let's say there were mitigating factors, such as emotional, cultural, or language difficulties which produced the low score. Or suppose this individual has been doing well in regular classes in high school, or obtained a "B" when he took a college class. To adequately interpret the IQ score, one needs to be able to weigh any factors which may have affected the score, and to also take into account the abilities that the individual has demonstrated in the real world. Generally "real world" performance is a better predictor than a test. Often tests are regarding as showing an individual's "true" ability or personality, whereas it makes much more sense to regard an individual's actual day-to-day behavior as reflecting true potential.
Personality tests are also frequently misinterpreted. The MMPI-2, a commonly used personality and psychopathology test, contains a "Schizophrenia" scale. But a high score on this test can mean many things other than schizophrenia. People often score high if they are feeling overwhelmed or have a history of substance abuse. But an untrained individual may incorrectly diagnose schizophrenia based on this one score. Tests may also be used to predict such things as "violence potential." But it is very difficult to predict whether a "potential" will ever manifest itself as violent behavior. The best predictor of future violence is a history of violence.
Tests are valuable tools with a strong potential for abuse. The best protection against test abuse is to seek assessment from a trained psychologist, and to ask for an explanation of test conclusions. An ethical professional reaches test conclusions based on test scores, usually from several tests, in combination with the client's history and the professional's clinical observations. A psychologist's extensive training aids him or her in giving proper weight to the various sources of information.