Validity is the term for how well a test measures what it has been designed to measure. Reliability refers to the ability to consistently find similar results when a test is repeatedly administered in similar conditions. High validity and high reliability will ensure that a test accurately determines what it claims to determine, and that its findings are not simply caused by random chance.

While a score of 100 indicates average intelligence on a modern IQ test, people who score 100 on these tests generally score much higher on older intelligence tests. Each generation since the 1930s has had increased scores on older tests, an effect studied by James Flynn.

Intelligence quotient (IQ) is defined as having a median of 100. This means that half of the people who take an IQ test will have below 100 and half will have above 100. The standard deviation of IQ is 15, such that two-thirds of the population fall between 85 and 115.

Construct validity refers to how closely a measurement accurately describes the underlying factor that it is trying to measure. An anxiety measurement would obviously not accurately describe participants' depressive symptoms.

External validity refers to the gerneralizability of your data. If there are only caucasian males in Stacy's study, her findings may not generalize to other groups, such as females or other races.

Reliability and validity are two important aspects to note when analyzing a set of data. Reliability is the degree to which a set of results can be repeated (ex. getting the same test score twice), while validity is the degree to which a set of data measures what it is designed to measure (ex. a depression test actually measures depressive symptomology).

Reliability refers to the ability of a study to produce similar findings if repeated by an independent team in similar conditions. Validity refers to a study's suitability for answering the question it was designed to investigate.

Content validity is the extent to which a scale holistically represents a given social construct, such as a personality trait or disorder. A scale that is high in content validity will represent a given social construct and all of its scientifically agreed upon assets—physical, emotional, behavioral, etc. A scale that is low in content validity will overlook one or some of these assets. For instance, a test trying to measure anxiety might ignore the physical aspects, such as sweating or high heart rate. This would be a test low in content validity.

An example of the halo effect would be if a handsome student is given bonus points on a test, though another student doesn't get bonus points for the same problem with the same answer.

Although some might have been tricked by the answer about America's obsession with celebrities, the halo effect is much more specific than idolizing a person because they are exceptional. Rather, the halo effect has to do with one positive trait spilling over to other traits. In this case, since the student is physically attractive, the teacher perceives them as more intelligent. Of course, this is a logical falllacy that can and does impede our judgement of people in our daily lives.

An example that would apply to celebrities is perceiving George Clooney to be a smart and kind idividual without knowing much about him or his personal life.

A test that is designed to predict a person's future success in a given area is an aptitude test. This measure has less so to do with a person's current skill set—though that is certainly relevant—and more to do with their talents, inclinations, and interests. A person's aptitude has to do with where a person has the most potential for growth.