What is intelligence? First of all, it is important to note that there is not a universal definition of intelligence. Yet, there is a surprisingly good agreement between most psychologists of what constitutes intelligence. The following terms are commonly used in the definition of intelligence: think abstractly, solve problems, identify relationships, learn quickly, memory functions, speed of mental processing, learn from experience, plan effectively, deal effectively with symbols.
The first of two classical theories of intelligence is developed by Charles Spearman (1904, 1927). Spearman claimed that test performance was mostly affected by one, general mental ability, called "g". In addition to this general factor, each test had some unique or specific variance, denoted by "s". Moreover, each "s" carried some unique variance that is due to a specific ability, and error variance. His theory is also referred to as two-factor theory, because it has two factors (g and s). However, the g-factor is the dominant one in his theory.
The second classical theory of intelligence is developed by the American psychologist L. L. Thurnstone. In contrast to Spearman, Thurnstone did not believe in a single, general underlying factor of intelligence. His theory is referred to as the multiple-factor theory. He extracted twelve factors of intelligence, of which he considered nine interpretable. The nine factors of intelligence are: (1) spatial; (2) perceptual; (3) numerical; (4) verbal; (5) memory; (6) words; (7) induction; (8) reasoning; (9) deduction.
The 'battle' between the first and second theory is also referred to as the one versus many argument. A compromise position is sought for by hierarchical models. In a hierarchical model, there are many separate abilities, but these are arranged in a hierarchy, with just one or a few dominant factors of intelligence on top of the hierarchy. Examples of such hierarchical models are developed by Catell (fluid versus crystallized intelligence), Vernon, and Caroll (three-stratum theory, with Spearman's "g" at the highest level, or stratum).
A topic on which there is much debate in the literature concerns the influences of heredity (nature) and environment (nurture). However, the question is no longer whether it is nature or nature. All scholars nowadays agree that intelligence results from an interaction between these two components. In addition, these two influences are not additive, but are related more in a sort of multiplicative relationship. That is, if one is completely negative, no matter how high the other is, intelligence will not develop. Another misconception is that nature traits are already present at birth, whereas nurture influences develop later. This is not true. Think, for example, about baldness. While this is not visible at birth already and does not manifest until midlife or later, it is already present at birth in the genes.
In sum, there are nine common features of intelligence tests: (1) individual administration; (2) it requires advanced training to administer the test; (3) broad range of ages and abilities; (4) establishment of rapport; (5) free-response formatted items; (6) immediate scoring of items; (7) test administration takes about one hour; (8) it allows for observation.
Although individual intelligence tests may be very different, they usually contain items covering the following nine categories: (1) vocabulary; (2) verbal relations; (3) information; (4) meaning, comprehension; (5) arithmetic; (6) short-term memory; (7) form patterns; (8) psychomotor; (9) matrices.
For many years, the Stanford-Binet test was the most popular method for measuring intelligence. However, two drawbacks of this test were (according to David Wechsler) that is was oriented on children only, and that it yielded only a single, general score. Hence, Wechsler developed the so-called Wechsler Adult Intelligence Scale (WAIS), which was first published in 1939. Since then, many revisions and new editions are published. Somewhat paradoxical, he later also developed the Wechsler Intelligence Scale for Children (WISC) for children aged 6-16 years, and the Wechsler Preschool and Primary Scale of Intelligence (WPPSI) for children aged 21-27 months. Even now, new versions are published using his name, even though Wechsler died in 1949.
Which one is better, the Stanford-Binet test or the Wechsler scales? This question remains unanswered. However, these two series nowadays certainly are much more similar than they were in the past.
In the early twenty-first century, professionals became increasingly concerned about the stigmatisation of these terms. An alternative term was found in intellectual disability. The definition of intellectual disability depends heavily on the concept of adaptive behaviour, that is, how well a person copes with ordinary life. Examples of adaptive behaviours are: feeding oneself, clothing (level 1), reading simple words (level 2), and taking the bus (level 3). The formal definition of intellectual disability yields three criteria that must be met for intellectual disability: (1) a significant subaverage intellectual functioning; (2) limitations in adaptive behaviour; (3) onset before age 18.
Group mental ability tests are mainly used in (a) elementary and secondary schools, with achievement tests; (b) predicting school success in college and graduate or professional school; (c) job selection or placement in military and business, and; (d) research.
Similar to the individually administered tests, there are a number of characteristics that are very common in group tests of mental ability. The eight common characteristics are: (1) obviously, the test can be administered to large groups in which there, theoretically, is no limit to the size of the group; (2) multiple-choice items, so that the tests are amenable to machine-scoring; (3) the content of the test is very similar to the individually administered tests; (4) fixed time limit and fixed number of items; (5) administration time is very bimodal distributed, administration often takes about one or three hours; (6) total score and several subscores; (7) very large research base for norming, equating, reliability and so on; (8) their primary purpose is to predict future success in school or on the job.
Six generalisations can be made about group mental ability tests. First, although there are huge differences in target groups and purpose of the tests, there is notable similarity in their content. Commonly, there are items on vocabulary, verbal relationships, reading, and so on. Second, total scores of the group mental ability tests are commonly very reliable, with internal consistency reliabilities around .95 and test-retest reliabilities around .90. Third, there is notable similarity in predictive validity of these tests. Fourth, a lack of differential validity is identified. Various combinations of subtests often do not yield higher validity coefficients than obtained with the total scores. Fifth, there are two statistical issues with these group tests: range restriction and imperfect reliability. Sixth, so far, there is no successful culture-free test to measure intelligence with a group test.
What is the difference between ability tests and achievement tests? It is best to think of it as a continuum, rather than considering these tests as rigidly distinct compartments. The point on which a test falls within the ability-achievement continuum depends on the degree of specific training. At the extreme right of the continuum (that is, specific achievement) fall tests that are highly dependent on specific training. For example, tests related to historical facts (civil war battles) or specialised skills (riding a bicycle). On the extreme left of the continuum (general ability) fall tests that require little dependence on specific training. Examples are solving puzzles, and detecting patterns. In the middle then, are traits such as reading comprehension and arithmetic problems.
Although there are huge differences between achievement tests, we identified five common features. First, most batteries -although referred to as "a" test- are a system of many interrelated tests. Second, the tests are oftentimes accompanied by a large amount of supplementary materials (detailed lists of objectives, interpretive booklets for students, teachers, parents, school administrators, computer-generated scoring reports, and so on). Third, there often are exemplary norming procedures and other research programs, with exhaustive technical manuals. Fourth, while achievements tests traditionally relied solely on multiple-choice items, they now often are accompanied by free-writing exercises and open-ended questions. Fifth, many achievement batteries depend on the same sources of information for their content, for example major textbook series and outlines provided by the National Assessment of Educational Progress.
There are both similarities and differences between tests of normal personality and clinical instruments. Four similarities are: (1) nature of items (simple), and response format (simple); (2) subdivisions for comprehensive and specific domain tests; (3) strategies for development; (4) threat of response sets and faking. Three differences are: (1) orientation; (2) administration setting; (3) manual and purpose.
One way of testing in clinical settings is by using an interview. First, we have to distinguish between the different types of interviews: unstructured, semistructured and structured. Note that these three types are not discrete categories, but that they fall on a continuum. At the one end, the unstructured (traditional) interview does not follow a certain pattern and consequently varies from one respondent to another, as well as from one examiner to another. At the other end of the spectrum, an structured interview uses a strict pattern. The same topics, with the same questions, are administered with each respondent. The semistructured interview falls between these two approaches. There are some standard questions, but there is also flexibility to tailor the interview to the individual respondent. In this section, we will mainly focus on the structured clinical interview.
The MMPI-2 is an extensive self-report inventory, consisting of 567 items. The MMPI was first published in 1942. The revision (second edition) appeared in 1989. Test administration usually takes about 60 to 90 minutes (120 for examinees with low reading levels or high distraction levels). This test is widely used. It is the most frequently used test by neuropsychologists and the second most frequently used test by clinical psychologists. The MMPI has its own language, customs, and rituals. For instance, a respondent could be described as: "24 code type with elevated F scale". An elevated F score may indicate, among others, severe pathology or "a cry for help".
A group of clinical instruments that yields their own category is the Behaviour Rating Scales (BRS). These scales are widely used to determine conditions such as attention disorders and assorted emotional problems. Two important features of these scales are: (1) someone else completes the rating, typically a teacher; (2) it lists specific behaviours, and the descriptors are short, usually one to three words.
Two pioneers in the field of career assessment are Edward K. Strong Jr. and G. Frederic Kuder. Broadly speaking, there are two traditional differences in approaches to career interest measurement: (1) origin of scales: criterion-keying versus broad areas; (2) item format: absolute versus relative level of interest.
Based on our examination of the three widely used career interest inventories, we derived the following five generalisations. First, career-related interest patterns of individuals seem to be quite reliable, at least from middle adolescence onward. Second, measures or career interest have a respectable degree of validity. It appears that different occupational groups tend to differ in their interest patterns, and that people tend to enter occupations that are consistent with their interests. Third, manuals of these tests provide little or no references to more modern psychometric techniques (such as IRT of differential item functioning). Fourth, career interest inventories are increasingly administered online. Fifth, not surprisingly, there is a positive relationship between interests and ability. Career interest testing must be accompanied by relevant information from the ability domain. Someone who wants to be a physician, but does not have the ability to successfully complete the requisite science course, is not suitable for the occupation.
Measurement of attitudes overlaps partly with public opinion polls. What is the difference between these? Clearly, there is no difference in the nature of the questions. Items from public opinion polls could easily be used in attitude scales and vice versa. The difference lies in the target for inference. Attitude measures aim to assess the attitude of an individual. Public opinion polls aim to assess the position of the group.
Add new contribution