Psychology, 5/e   Wortman, Loftus & Weaver
	Online Learning Center

CONCEPT I: Constructing and Evaluating Tests

Tests are widely used in our society for a wide variety of purposes. Tests can tell us many things about individuals and groups, but must be used with caution because their results can and do affect people’s lives.

In the process called standardization, uniform procedures for giving and scoring a test are developed. A test is not very meaningful until we compare it to its norm, the range of scores attained on the test by other similar people. Before a test is put into general use, it is given to a well-defined and representative group of people to which other test results can be compared so that appropriate norms can be developed. These norms provide a kind of benchmark.

A good test must have reliability--that is, it must consistently yield the same results. Test-retest reliability is determined by administering the same test to the same individuals on more than one occasion. If reliability is high, the individuals’ scores on both occasions should be similar. Split-halves reliability, also called internal consistency, is often used when retesting is impossible or undesirable. Using the split-halves method, the test items on a particular test are randomly divided into two groups and performances on the two halves are compared.

Validity reveals whether the test is measuring what it is intended to measure. Criterion validity refers to the relationship between a person’s test score and some other criterion. Predictive validity compares test performance to performance on some future task.

CONCEPT II: Measuring Intelligence

Intelligence is the set of capacities required to acquire and retain knowledge and to understand concepts and relationships which are used in effective problem-solving and communication. Two approaches to defining intelligence are factor analysis and the information-processing approach. Researchers have tried to identify various factors, the components of intelligence. Charles Spearman has proposed that all the items on an intelligence test measure a general, or "g," factor. Raymond Cattell proposed that there are two g factors, gf (fluid intelligence), which measure reasoning and understanding and was thought to be mostly inherited, and gc (crystallized intelligence), which is built from experience with the environment. Fluid intelligence declines in later adulthood, but crystallized intelligence keeps increasing until about age 60. Arthur Jensen has observed that reaction time on easy inform ation-processing tasks correlates strongly with traditional measures of g. Jensen has proposed that this correlation reflects certain neurologically based components of intelligence. Louis Thurstone identified seven factors in intelligence, and J. P. Guilford identified 150. There is no general agreement about the correct number of independent factors in intelligence.

The information-processing view of intelligence is based on an understanding of the cognitive activities involved in intelligence. Robert Sternberg, for example, believes that psychologists who study intelligence should try to understand the various mental operations that typify a person’s ability to think. John Carrol’s studies suggest a three-tiered hierarchical model of intelligence, with g at the top, learning and memory at the second level, and specific skills which meet environmental demands at the bottom.

Borrowing from his cousin Charles Darwin’s theory of evolution, Francis Galton believed that psychological traits such as intelligence and moral character were inherited. He advocated eugenics, the selective breeding to improve the human species, and he developed an intelligence test to select those people of highest ability. Galton’s test measured reaction time, sensory acuity, physical energy, and head size, none of which was found later to be related to intelligence.

In the early 1900s, Alfred Binet developed the first tests of mental ability (intelligence tests) designed to predict who would be successful in school. He reasoned that since older children could typically answer more difficult questions than younger children, individual children could be characterized by mental age. The IQ, or intelligence quotient, is the ratio of a child’s mental age as determined by Binet’s test to the child’s chronological age (how old the child is) multiplied by 100.

Although Binet believed that intelligence was not a fixed trait, Louis Terman, who brought Binet’s test to the United States, and several other influential Americans did. Consequently, this view led to several political decisions, such as sterilization of retarded people and restriction of immigrants based on IQ, which we now view as inappropriate.

The revision of Binet’s original intelligence test currently in use is the Stanford-Binet test. This test contains items from a number of verbal and performance subtests arranged in order of difficulty. A child’s mental age is compared to his or her chronological age multiplied by 100 to yield IQ. Today, IQ is no longer calculated by the mental age-chronological age ratio. Instead, a score is compared to a norm group to see how far the score deviates from the norm (average) for children the same age. Now, the Stanford-Binet has a mean score of 100 and a standard deviation of 15. Instead of calling the scores IQs, the appropriate term now is standard score.

The most frequently used individual intelligence tests are the Wechsler Preschool and Primary Scale of Intelligence (WPPSI). The Wechsler scales report separate scores for each of the several subtests and yield two intelligence scores (verbal and performance), unlike the Stanford-Binet, which yields only one. The Wechsler tests are less biased toward verbal skills, and they use the same items for subjects within each scale. The Wechsler tests encourage the view that intelligence is composed of a number of different abilities.

Paper-and-pencil tests of intelligence can be administered to large groups of people. One such test is the Scholastic Aptitude Test (SAT). Although group tests are less expensive and easier to administer and score, they are not as accurate in predicting school success as are individual tests. Whether or not coaching can improve a score on the SAT remains a controversial issue, but it is doubtful that coaching can produce large improvements in test performance. Critics of the SAT have argued that success on the test is more a product of upper-middle-class background than of scholastic aptitude. Its defenders have pointed out that both the verbal and mathematical subtests of the SAT are good predictors of college success, especially when combined with high school performance.

Howard Gardner believes traditional IQ tests measure only language ability, mathematical-logical reasoning, and spatial-perceptual skills. He believes that intelligence also is composed of musical ability, bodily ability, intrapersonal ability, and interpersonal ability. In similar fashion, Robert Sternberg believes that we should expand our current understanding of intelligence to include the ability to adapt to unexpected situations, to have sudden insights when solving problems, to learn from context, and to perform different tasks at the same time.

Of the total population, about 68% score in the "normal" range (plus or minus 1 standard deviation) on intelligence tests (between 85 and 115); 95% fall between 70 and 130; and almost 98% fall between 55 and 145.

A person who has a significantly below average intelligence and difficulties with normal functioning is usually classified as mentally retarded. Levels of retardation correspond to IQ test scores; 70 to 50 is considered mild retardation; 49 to 35, moderate retardation; 34 to 20, severe retardation; and less than 20, profound retardation. These categories are not absolute, however, and depend somewhat on the individual’s adaptive skills. Since social and perceptual limitations can cause a person of normal intelligence to seem retarded, careful screening must be done for an accurate diagnosis.

Among the many misconceptions about children who are intellectually brilliant are beliefs that their brilliance fades with age, and that they are unhealthy, mentally unstable, and socially inept. In 1921 Terman began studying 1,500 child geniuses and found that such beliefs about the mentally gifted were not valid. In fact, he found that gifted children often excel in all aspects of life, both intellectual and social. Terman also found that a good environment has a favorable effect on intelligence. Researchers have found that education geared to the gifted is beneficial. Such research points out that giftedness benefits from challenges and develops best in an enriched environment. Still puzzling is the finding that gifted women still achieve and earn less than gifted men.

Intelligence is the result of the interaction of inherited factors and environmental factors. Although the genetic component of intelligence cannot be measured directly, there is considerable agreement that, in general, about 50% of the variability in intelligence is inherited, and about 50% results from environmental differences. Thus, the heritability factor, or the estimate of the genetic contribution to a given trait, for intelligence is about 50%. The heritability factor does not tell us anything about a given individual’s intelligence, nor is it uninfluenced by the environment in which it is estimated. High achievers tend to come from homes with supportive parents, although new research suggests that each child in a family experiences an environment quite different from those experienced by the other children.

CONCEPT III: Assessing Personality

There are many different kinds of personality tests. The type of test used often depends on the theoretical approach of the tester. Projective tests of personality ask the participant to interpret ambiguous material and in so doing reveal his or her own conflicts and emotions. Self-reports attempt to measure how many of a particular group of traits a person has. One type of objective test is developed by using statistics to disclose which test items best measure these basic traits. The resulting tests are validated to make sure they work properly. Other tests like the Minnesota Multiphasic Personality Inventory (MMPI) are developed empirically by administering the test to various groups of participants and comparing responses. The MMPI is the most widely used test in outpatient clinics and it is composed of 550 true-false items that may pertain to a person’s self-image. An individual’s pattern of answers is compared to typi cal answers given by various groups of diagnosed psychiatric patients. A high score on one set of items (or scale) could indicate personal characteristics similar to those typical of the group from which the scale was drawn (for example, victims of depression). The ten clinical scales (such as depression) are supplemented with three validity scales, the Lie scale (which indicates lying), the F scale (which picks out highly unusual answers), and the K scale (which reveals the participant’s evasiveness in answering the questions). The MMPI is useful for identifying and diagnosing psychopathology, but not for measuring normal personality. The MMPI has been revised and published as the MMPI-2, which contains 567 items and several new scales.

Perhaps the best-known projective test, the Rorschach Inkblot Test asks participants to report what they see in each of 10 inkblots. Their responses are scored according to the part of the inkblot they interpret, the form, color, or shading of the blot they emphasize, and what they claim to see. Participant response can vary depending on how and by whom the test is administered. The test has been criticized for not being a reliable measure of personality, but others argue that it is very useful in gaining insight into a person’s mind. The Thematic Apperception Test (TAT) requires the participant to tell a coherent story about each of several pictures of people doing different things. Originally intended to reveal the impact of a person’s internal needs and environmental factors, it is now often used as a general personality test in both clinical and research settings. It has many of the same advantages and disadvantages as the Rorschach.

Tests can be used to help people decide which careers might be best for them. Interest tests are useful in educational and vocational counseling. The Strong-Campbell Interest Inventory (SCII) presents 325 items. The participant reports whether he or she likes, dislikes, or is indifferent about each item. An SCII report consists of three types of information: the particular occupations that have interest patterns similar to that of the test taker (the Occupational Scale), the areas of interest the person scores high on (the Basic Interest Scale), and adjectives that describe the person’s general styles of thinking and acting (the General Occupational Themes Scale).

Screening tests are designed to identify applicants who are most likely to perform well on a particular type of job. Screening tests are sometimes administered to measure specific aptitudes (such as finger dexterity) or personality characteristics (such as assertiveness). Employers must be sure that their screening tests are not unfairly biased against certain groups.

CONCEPT IV: The Ethics of Testing

Because tests are used extensively and their results affect people’s lives, they must be used ethically. They must only be used when valid--when they measure accurately what they are intended to measure. They cannot be used when they unfairly discriminate against minority groups. Since tests do not yield certain information, they should always be used in combination with other information when important decisions must be made, and people who score low should be given a chance to retake their tests. Finally, the results of tests should be kept confidential and, whenever possible, test takers should be able to see how their answers were scored.

CONCEPT V: The Bell Curve: Are There Racial Differences in IQ?

Great controversy has arisen in recent years over how much of a role genetics plays in intelligence, especially in racial differences in IQ scores. English psychologist Sir Cyril Burt observed strong influences of heritability on IQ in White populations, and Arthur Jensen sought to determine whether average IQ differences between Whites and Blacks might also be related to inherited factors. Jensen was immediately criticized for his studies, although he never actually asserted that Whites score higher on IQ tests because of genetics. The recent book The Bell Curve (1994) by Richard Herrnstein and Charles Murray has again ignited heated debate on the subject. This book argues that we can not ignore racial differences in intelligence when considering socioeconomic disparity between races. Other psychologists argue that intelligence accounts for only a small amount of variance in socioeconomic differences, and that the environment in which one grow s up is probably much more influential on one’s IQ score.

Copyright ©1999 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use and Privacy Policy.
McGraw-Hill Higher Education is one of the many fine businesses of The McGraw-Hill Companies.
For further information about this site contact mhhe_webmaster@mcgraw-hill.com.