Representation, Norms, and Validity - AP Psychology

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.

An achievement test is a test that is designed to assess a person's current skill set. An achievement test often contrasts with an aptitude test, which has more to do with a person's consistent traits that will predict their future success in particular areas. Achievement tests are more often administered in schools to test a student's given mastery of particular material. This very quiz is an example of an achievement test.

"Norms," with regard to research and testing, refers to the percentile ranks of a large and representative sample of a particular demographic for which the test was designed. This is not to be confused with societal norms, which have to do with the rules people adhere to in society.

A standardized score reflects a person's status relative to the average of all scores on the given evaluation. That is, the standard score reflects a person's relationship with the mean of all scores. If the average of all scores is 70, and a person scores 90, the person's standardized score is positive 20. If the person scores 20 below the mean, their standardized score would be negative 20.

Someone who is high in intrapersonal intelligence fuctions at their highest level while alone. This can often translate into high-self esteem, since they provide for their own individual needs instead of relying on other people for support. They process their lives during their alone time, reflecting on succcesses and mistakes. This contrasts with people who have high interpersonal intelligence, who may get these needs met from communicating with other people.

When a researcher tries to take measurements, they are doing so in the most precise and controlled manner possible. However, there will be some moments when this precision and control is not always possible, and a chance variation may occur. The researcher will always try and reduce the amount of chance variations to the extent that they can.

Descriptive statistics are primarily concerned with numbers. Its generalizations are primarily mathematical. Inferential statistics, however, are more concerned with the generalizations that can be made from a given data set. One must be sure, though, that the data set is a represenative sample of a given population. Inferential statistics allows for researchers to make statements such as "50-year-old males prefer to eat hamburgers for breakfast" given their numerical data.

In a numerical data sheet, numbers can vary greatly. The measure of how much variation (or dispersion) there is is called the measure of variation.

A valid test will measure what it is intended to measure. For instance, a test designed to help diagnose depression must ask questions that pertain to depressive disorders and offer answer options that are relevant to attaining a diagnosis. If a test designed to diagnose depressions accurately assesses patients, then it is said to have high validity. If a test designed to diagnose depressive disorders instead presents positive results for a random sample of individuals, then it must have low validity; it fails to perform its designed function.

External validity looks at the extent to which a study's findings can be generalized to certain populations. Studies that have large sample sizes and diverse participants have greater external validity.

Validity and reliability are both important terms for experimental design and are both markers of a good study. A study is valid if it measures what it is supposed to measure and reliable if it produces similar results every time it is conducted. If it is not valid, then it cannot answer the hypothesis or be applied to the real world; furthermore, if it is not reliable, then the results cannot be trusted.