lecture 5 - personality and intelligence Flashcards

Question

allport's search for traits

Answer 1

Gordon Allport (1897–1967) was one of the first psychologists to search systematically for a basic core of personality traits. He began by identifying all the words in an unabridged dictionary of the English language that described aspects of personality (Allport and Odbert, 1936). He found around 18,000 words, which he then further analysed for those that described only stable personality characteristics. He eliminated words that represented temporary states, such as ‘flustered’, or evaluations, such as ‘admirable’. This still left him with over 4,000 words. Allport was interested in learning how many traits are needed to describe personality and exactly what these traits may be. For example, many of those 4,000 words, such as ‘shy’ and ‘bashful’, are synonyms. Although each synonym presumably makes some sort of distinction about a trait, a group of synonyms together might be used to describe the same underlying trait. Many trait theorists believe that the most basic set of personality traits ranges from three to sixteen traits. Allport’s research stimulated other psychologists to think about personality in terms of traits or dispositions. In fact, most modern trait theories can be traced to Allport’s earlier theoretical work. Like Allport, modern trait theorists maintain that only when we know how to describe an individual’s personality will we be able to explain it.

Answer 2

Factor analysis is an important means of defining intelligence (see Chapter 11). Factor analysis identifies variables that tend to be correlated. To use factor analysis to study personality, researchers must observe the behaviour of a large number of people. Usually, the observations are limited to responses to questions on paper-and-pencil tests, but occasionally, investigators observe people’s behaviour in semi-natural situations. Statistical procedures then permit investigators to determine which items a given person tends to answer in the same way; they can then infer the existence of common factors. For example, a shy person would tend to say no to statements such as ‘I attend parties as frequently as I can’ or ‘When I enter a room full of people, I like to be noticed.’ In contrast, outgoing people would tend to say ‘yes’ to these statements. To the degree that people possess orderly personality traits, they tend to answer certain clusters of questions in particular ways. Raymond Cattell (1905–98) began his search for a relatively small number of basic personality traits with Allport and Odbert’s (1936) list of adjectives. In addition, he collected data on people’s personality characteristics from interviews, records describing their life histories, and from observing how people behave in particular situations. From this list, Cattell began to construct preliminary versions of a questionnaire called the 16PF. Then, using factor analysis, he analysed responses from thousands of people to whom the inventory had been administered. Eventually, he identified 16 personality factors. Cattell referred to these 16 traits as source traits because, in his view, they are the cornerstones upon which personality is built: they are the primary factors underlying observable behaviour. He called groups of similar types of observable behaviour surface traits; he included such traits as kindness, honesty and friendliness because they are visible to others. They represent the surface of personality and spring forth from source traits, which lie deeper within the personality. Figure 14.2 illustrates a personality profile of a hypothetical individual rated on Cattell’s 16 factors. The factors are listed in order of importance, from top to bottom.

Answer 3

Hans Eysenck (1916–1997) also used factor analysis to devise his theory of personality (Eysenck, 1970; Eysenck and Eysenck, 1985). His research identified three important factors: extraversion, neuroticism and psychoticism. These factors are bipolar dimensions. Extraversion is the opposite of introversion, neuroticism is the opposite of emotional stability, and psychoticism is the opposite of self-control. Extraversion refers to an outgoing nature and a high level of activity; introversion refers to a nature that shuns crowds and prefers solitary activities. Neuroticism refers to a nature full of anxiety, worries and guilt; emotional stability refers to a nature that is relaxed and at peace with itself. Psychoticism refers to an aggressive, egocentric and antisocial nature; self-control refers to a kind and considerate nature, obedient of rules and laws. Eysenck’s use of the term ‘psychoticism’ is different from its use by most clinical psychologists; his term refers to antisocial tendencies and not to a mental illness. A person at the extreme end of the distribution of psychoticism would receive the diagnosis of antisocial personality disorder. According to Eysenck, the most important aspects of a person’s temperament are determined by the combination of the three dimensions of extraversion, neuroticism and psychoticism, just as colours are produced by the combinations of the three dimensions of hue, saturation and brightness. More than most other trait theorists, Eysenck emphasises the biological nature of personality (Eysenck, 1991). For example, consider the introversion–extraversion dimension, which is biologically based, according to Eysenck, on an optimum arousal level of the brain. Eysenck believes that the functioning of a neural system located in the brain stem produces different levels of arousal of the cerebral cortex. Introverts have relatively high levels of cortical excitation, whereas extraverts have relatively low levels. Thus, in order to maintain the optimum arousal level, the extravert requires more external stimulation than does the introvert. The extravert seeks stimulation from external sources by interacting with others or by pursuing novel and highly stimulating experiences. The introvert avoids external stimulation in order to maintain their lower arousal level at an optimum state. Different states of arousal are hypothesised to lead to different values of the extraversion trait for different people. This hypothesis is reviewed in the section on the biological basis of personality. Most trait theorists accept the existence of Eysenck’s three factors because they have emerged in factor analyses performed by many different researchers; these appear, in fact, to have the highest validity of all proposed personality factors (Kline, 1993)

Answer 4

Languages reflect the observations of a culture; that is, people invent words to describe distinctions they notice. An analysis of such distinctions by Tupes and Christal (1961), replicated by Norman (1963), led to the most widely accepted model of personality traits: the five-factor model (FFM); McCrae and Costa, 1985, 1987, 1990). The FFM proposes that personality is composed of the following five primary dimensions: 1Openness 2Conscientiousness 3Extraversion 4Agreeableness 5Neuroticism which you can remember using the handy mnemonic: OCEAN. These factors can be measured by the Neuroticism, Extraversion and Openness Personality Inventory, or NEO-PI, which consists of 181 items that potentially describe the person being evaluated (McCrae and Costa, 1990). Studies have shown that people’s assessment of their own personality agrees well with ratings by spouses and those who know them. The test items are brief sentences, such as ‘I really like most people I meet’ or (for ratings by someone else) ‘She has a very active imagination’. The person taking the test rates the accuracy of each item on a scale of 1 to 5, from strong disagreement to strong agreement. The scores on each of the five factors consist of the sums of the answers to different sets of items. McCrae et al (1986) validated the FFM through the factor analysis of a list of adjectives contained in a test called the California Q-Set. This test consists of 100 brief descriptions (such as ‘irritable’, ‘cheerful’, ‘arouses liking’ and ‘productive’). The items were provided by many psychologists and psychiatrists who found the words useful in describing people’s personality characteristics. Thus, the words are not restricted to a particular theoretical orientation. McCrae and his colleagues found that factor analysis yielded the same five factors as the analysis based on everyday language: neuroticism, extraversion, openness, agreeableness and conscientiousness. The FFM is regarded by most personality psychologists as a fairly robust model of personality (Magai and McFadden, 1995) although it is a model that is atheoretical and descriptive rather than explanatory (Hampson, 2012). The Big Five Personality Inventory has been used to investigate various relationships between personality and other psychological variables, as you will see in this chapter and throughout the text. When a study investigates conscientiousness or openness to experience, for example, it is using the FFM model. For example, Zhao and Seibert (2006) investigated the differences between the personalities of managers and entrepreneurs. In a meta-analysis, they found that entrepreneurs scored more highly on the dimension of conscientiousness and openness to experience but scored lower on neuroticism and agreeableness. There was no difference between the two groups for extraversion

Answer 5

In 2002, Paulhus & Williams published a study in which they described three traits that reflected the darker aspects of human nature, the so-called Dark Triad. These were: narcissism defined as ‘the pursuit of gratification from vanity or egotistic admiration of one’s own attributes’; Machiavellianism, defined as ‘a duplicitous interpersonal style, a cynical disregard for morality, and a focus on self-interest and personal gain’; and psychopathy, defined as ‘antisocial behaviour, diminished empathy and remorse, and disinhibited or bold behaviour’ (Muris et al, 2017). A fourth was later added, sadism, resulting in a dark tetrad. These traits have been associated with a variety of (largely) negative behaviours from cheating and antisocial online behaviour (Moor & Anderson, 2019) to the use of aggressive humour (see Martin, 2020). Although they are presented as individual traits they are related. They are more commonly seen in men than women (Muris et al, 2017) and one – psychopathy – may be more predictive of unpleasant or negative behaviour than the others (Moor & Anderson, 2019; Muris et al, 2017). All three traits are negatively correlated with agreeableness and honesty/humility (Muris et al, 2017). Some research suggests that people can surmise a person’s level of dark tetrad trait from their facial expression. Shiramizu et al (2019), for example, found that people were able to identify high narcissism men and women at levels better than chance from photographs. Men showing high levels of psychopathy could be identified, too. Participants were unable to identify high Machievellianism types and identified high psychopathy women less often than expected by chance

Answer 6

Some longitudinal studies of personality show remarkable stability in personality factors (especially, extraversion). Cross-sectional studies, however, show less stability, not surprisingly, perhaps, because cross-sectional studies include people who differ in age (and, therefore, share different cultural influences). A study of the personalities of 163 men over 45 years found that neuroticism, extraversion and openness were positively correlated throughout the 45 years and that the traits remained relatively stable (Soldz and Vaillant, 1999). Overall, evidence suggests that we are capable of personality change and this is in a positive direction, towards greater maturity (Bleidorn et al, 2021). The time at which we show greatest personality change is young adulthood, but the change depends on our circumstances Some studies have suggested that some personality traits are more fluid than others. McCrae and Costa and their international team of researchers looked at changes in the Big Five personality traits from the age of 14 years to 30 years in a sample of Germans, British, Spaniards, Czechs and Turks (N = 5,085) (McCrae et al, 2000). They found that neuroticism, extraversion and openness to experience decreased from 14 to 30 years but that agreeableness and conscientiousness increased. Similar trends were found after 30 years, but the changes were not as pronounced. A similar pattern – some decline, some stability, but largely some change – has also been noted in a meta-analysis of longitudinal studies. Roberts et al (2006) found that social dominance (a feature of extraversion), conscientiousness and emotional stability increase between the ages of 20 and 40, whereas social vitality and openness increase during adolescence, then decline to old age. Smits et al (2011) asked first-year students between 1982 and 2007 to complete the Big Five. They found small increases in extraversion, agreeableness and conscientiousness and small decreases in neuroticism. No changes were found for openness to experience. A lack of stability in personality traits has also been reported in a longitudinal study of 174 people aged 77 (Harris et al, 2016). Harris et al were able to use data collected in 1947 when the individuals were adolescents. Of the large sample (N = 1, 208), a portion of the sample survived and agreed to take part in a follow-up study in 2012. In 1947, six personality traits were measured by the participant’s teachers. In 2012, the same traits were measured. People who knew the participants well were asked to judge their personality. While there was little stability found in any of the six traits nor in the general trait of dependability, when some other effects were excluded, mood stability and conscientiousness did show evidence of stability from adolescence to old age.

Answer 7

Objective personality tests are similar in structure to classroom tests. Most contain multiple-choice and true/false items, although some allow the person taking the test to indicate the extent to which they agree or disagree with an item. The responses that subjects can make on objective tests are constrained by the test design. The questions asked are unambiguous, and explicit rules for scoring the subjects’ responses can be specified in advance. Examples include the Eysenck Personality Inventory and the NEO-PI, described earlier. One of the oldest and most widely used objective tests of personality is the Minnesota Multiphasic Personality Inventory (MMPI), devised by Hathaway and McKinley in 1939. The original purpose for developing the test was to produce an objective, reliable method for identifying various personality traits that were related to a person’s mental health. Improvement in people’s scores over the course of treatment would indicate that the treatment was successful. In devising this test, Hathaway and McKinley wrote 504 true/false items and administered the test to several groups of people in mental institutions in Minnesota who had been diagnosed as having certain psychological disorders. These diagnoses had been arrived at through psychiatric interviews with the patients. Such interviews are expensive, so a simple paper-and-pencil test that accomplished the same result would be valuable. The control group consisted of relatives and friends of the patients, who were tested when they came to visit them. (Whether these people constituted the best possible group of normal participants is questionable.) The responses were analysed empirically, and the questions that correlated with various diagnostic labels were included in various scales For example, if people who had been diagnosed as paranoid tended to say true to ‘I believe I am being plotted against’, this statement would become part of the paranoia scale. The current revised version of the MMPI, the MMPI-2, has norms based on a sample of people that is much more representative ethnically and geographically than the original sample (Graham, 1990). It includes 550 questions, grouped into ten clinical scales and four validity scales. A particular item can be used on more than one scale. For example, both people who are depressed and those who are hypochondriacal tend to agree that they have gastrointestinal problems. The clinical scales include a number of diagnostic terms traditionally used to label psychiatric patients, such as hypochondriasis, depression or paranoia. The four validity scales were devised to provide the tester with some assurance that subjects are answering questions reliably and accurately and that they can read the questions and pay attention to them. The ‘?’ scale (‘cannot say’) is simply the number of questions not answered. A high score on this scale indicates either that the person finds some questions irrelevant or that the person is evading issues that they find painful. The L scale (lie) contains items such as ‘I do not read every editorial in the newspaper every day’ and ‘My table manners are not quite as good at home as when I am out in company’. A person who disagrees with questions like these is almost certainly not telling the truth. A high score on the L scale suggests the need for caution in interpreting other scales and also reveals something about the participant’s personality. In particular, people who score high on this scale tend to be rather naive; more sophisticated people realise that no one is perfect and do not try to make themselves appear to be so. The F scale (frequency) consists of items that are answered one way by at least 90 per cent of the normal population. The usual responses are ‘false’ to items such as ‘I can easily make other people afraid of me, and sometimes do it for the fun of it’, and ‘true’ to items such as ‘I am liked by most people who know me’. A high score on this scale indicates carelessness, poor reading ability or very unusual personality traits. The K scale (defensiveness) was devised to identify people who are trying to hide their feelings to guard against internal conflicts that might cause them emotional distress. A person receives a high value on the K scale by answering ‘false’ to statements such as ‘Criticism or scolding hurts me terribly’ and ‘At periods, my mind seems to work more slowly than usual’. People who score very low on this scale tend to be in need of help or to be unusually immune to criticism and social influences. Some psychologists argue that validity scales are useless or even harmful in most testing situations. For example, consider the following item: ‘Before voting, I thoroughly investigate the qualifications of all candidates.’ According to Crowne and Marlowe (1964), anyone who answers ‘yes’ to such a question has to be lying. But as McCrae and Costa (1990) note, people taking tests do not necessarily respond passively to each item, taking it at face value. Instead, their response is based on their interpretation of what they think the question means. They suggest (p. 40) that most people will say to themselves: Surely these psychologists did not mean to ask if I actually study the voting records of every single political candidate, from President to dogcatcher. No one does, so that would be a stupid question to ask. What they must have meant to ask was whether I am a concerned citizen who takes voting seriously. Since I am and I do, I guess I should answer yes There is evidence to support McCrae and Costa’s suggestion. When psychologists calculate a person’s score on the MMPI, they usually apply a correction factor derived from the validity scales. Several studies have shown that the application of the correction factors to the scores of normal subjects actually reduces the validity of these scores. McCrae and Costa suggest that when the MMPI is administered to normal subjects for research purposes, such corrections should not be made. However, validity scales may be useful in situations in which subjects may be motivated to lie (e.g. when a personality test is used to screen job applicants) or in cases in which the test is being used clinically to evaluate the possibility of mental illness or personality disorder. As well as being used in clinical assessment, the MMPI has been employed extensively in personality research, and a number of other tests, including the California Psychological Inventory and the Taylor Manifest Anxiety Scale, are based on it. However, the MMPI has its critics. As we saw earlier, the five-factor model of personality has received considerable support. Some of its advocates have noted that the MMPI misses some of the dimensions measured by the NEO-PI, which includes tests of neuroticism, extraversion, openness, agreeableness and conscientiousness (Johnson et al., 1984). Thus, these factors will be missed by a clinician or researcher who relies only on the MMPI. For this reason, many researchers, especially those interested in the psychobiology of personality, no longer use the MMPI

Answer 8

Projective tests of personality are different in form from objective ones and are derived from psychodynamic theories of personality. Psychoanalytically oriented psychologists believe that behaviour is determined by unconscious processes more than by conscious ones. Thus, they believe that a test that asks straightforward questions is unlikely to tap the real roots of an individual’s personality characteristics. Projective tests are designed to be ambiguous so that the person’s answers will be more revealing than simple agreement or disagreement with statements provided by objective tests. The assumption of projective tests is that an individual will ‘project’ his or her personality into the ambiguous situation and thus make responses that give clues to this personality. In addition, the ambiguity of the test makes it unlikely that subjects will have preconceived notions about which answers are socially desirable. Thus, it will be difficult for a subject to give biased answers in an attempt to look better (or worse) than he or she actually is.

Answer 9

One of the oldest projective tests of personality is the Rorschach Inkblot Test, published in 1921 by Hermann Rorschach, a Swiss psychiatrist. The Rorschach Inkblot Test consists of 10 pictures of inkblots, originally made by spilling ink on a piece of paper that was subsequently folded in half, producing an image that is symmetrical in relation to the line of the fold. Five of the inkblots are black and white, and five are colour. The participant is shown each card and asked to describe what it looks like. Then the cards are shown again, and the participant is asked to point out the features they used to determine what was seen. The responses and the nature of the features the participant uses to make them are scored on several dimensions. In the following example described by Pervin (1975), a person’s response to a particular inkblot might be: ‘Two bears with their paws touching one another playing a game or could be they are fighting and the red is the blood from the fighting.’ The classification of this response, also described by Pervin (p. 37), would be: large detail of the blot was used, good form was used, movement was noted, colour was used in the response about blood, an animal was seen, and a popular response (two bears) was made. A possible interpretation of the response might be as follows: Although the interpretation of people’s responses to the Rorschach Inkblot Test was originally based on psychoanalytical theory, many investigators have used it in an empirical fashion. That is, a variety of different scoring methods have been devised, and the scores obtained by these methods have been correlated with clinical diagnoses, as investigators have done with people’s scores on the MMPI. However, the validity of these scoring techniques and the validity of the test in general is questionable (Groth-Marnat, 1997). Individual starts off with popular response and animals expressing playful, ‘childish’ behaviour. Response is then given in terms of hostile act with accompanying inquiry. Pure colour response and blood content suggest he may have difficulty controlling his response to the environment. Is a playful, childlike exterior used by him to disguise hostile, destructive feelings that threaten to break out in his dealings with the environment?

Answer 10

Another popular projective test, the Thematic Apperception Test (TAT), was developed in 1938 by the American psychologists Henry Murray and C.D. Morgan to measure various psychological needs. People are shown a picture of a very ambiguous situation and are asked to tell a story about what is happening in the picture, explaining the situation, what led up to it, what the characters are thinking and saying, and what the final outcome will be. Presumably, the participants will ‘project’ themselves into the scene, and their stories will reflect their own needs. As you might imagine, scoring is difficult and requires a great deal of practice and skill. The tester attempts to infer the psychological needs expressed in the stories. Consider the responses of one woman to several TAT cards, along with a clinician’s interpretation of these responses (Phares, 1979, p. 273). The questions asked by the examiner are in parentheses. Card 3BM. Looks like a little boy crying for something he cannot have. (Why is he crying?) Probably because he cannot go somewhere. (How will it turn out?) Probably sit there and sob himself to sleep. Card 3GF. Looks like her boyfriend might have let her down. She hurt his feelings. He’s closed the door on her. (What did he say?) I don’t know. Card 10. Looks like there’s sorrow here. Grieving about something. (About what?) Looks like maybe one of the children’s passed away. Interpretation. The TAT produced responses that were uniformly indicative of unhappiness, threat, misfortune, a lack of control over environmental forces. None of the test responses was indicative of satisfaction, happy endings, etc . . . In summary, the test results point to an individual who is anxious and, at the same time, depressed. The pattern of responses in this case is quite consistent; few people would disagree with the conclusion that the woman is sad and depressed. However, not all people provide such clear-cut responses. As you might expect, interpreting differences in the stories of people who are relatively well adjusted is much more difficult. As a result, distinguishing among people with different but normal personality traits is hard. One major problem with the TAT is in quantifying responses, such as the ones above. Often, responses are analysed qualitatively which makes assessing the reliability of the test difficult. Others have argued that subjecting the test to quantitative rigorous examination defeats the object of using the test which is to help guide a clinician’s assessment of a patient’s personality. However, even here, there are problems in that there is little agreement between clinicians regarding the assessment of the individual’s responses on the TAT (Groth-Marnat, 1997).

Answer 11

In general, if people do well academically or succeed at tasks that involve their heads rather than their hands, we consider them to be intelligent. If a politician makes a useful policy decision, we call it an intelligent decision. If an author writes an erudite book on an arcane subject, we might describe them as having written an intelligent appraisal. But if asked to give a precise definition of intelligence, psychologists – in common with non-scientists – come slightly unstuck. Sternberg and Detterman (1986) asked a dozen theorists to provide definitions of intelligence and received a dozen different descriptions. According to one of psychology’s historians, writing in the 1920s, intelligence has come to represent whatever intelligence tests measure (Boring, 1923). In general, however, psychologists broadly agree that the term intelligence describes a person’s ability to learn and remember information, to recognise concepts and their relations, and to apply the information to their own behaviour in an adaptive way (Neisser et al, 1996a). Where they diverge is in describing the nature of intelligence and how it works. For example, some psychologists argue that there is a general factor called intelligence but no different subtypes of intelligence; others argue intelligence is a series of abilities; yet others adopt a combinative approach arguing that there is general intelligence but there are also specific abilities. The number of these abilities depends on the theory one examines

Answer 12

Most theories of intelligence are based on the analysis of performance on tests that seek to measure specific abilities such as non-verbal and verbal intellectual competence. Much of the debate in the psychology of intelligence has focused on whether there is a single intelligence or multiple intelligences. Is our intellectual ability a unitary factor or is it made up of several different abilities? Are these abilities, if they do exist, separate from each other or are they related? Intelligence tests yield a single number, usually called an intelligence quotient (IQ) score, although this does not itself mean that intelligence is a single, general characteristic. Some investigators have suggested that certain intellectual abilities are completely independent of one another. For example, a person can be excellent at spatial reasoning but poor at solving verbal analogies. But psychologists disagree over whether specific abilities are totally independent or whether one general factor influences all abilities. The next sections consider some of these disagreements and influential theories of intelligence.

Answer 13

Charles Spearman (1927) proposed that an individual’s performance on a test of intellectual ability is determined by two factors: the g factor, which is a general factor, and the s factor, which is a factor specific to a particular test. Spearman did not call his g factor ‘intelligence’; he considered the term too vague. He defined the g factor as comprising three ‘qualitative principles of cognition’: apprehension of experience, eduction of relations and eduction of correlates. A common task on tests of intellectual abilities – solving analogies – requires all three principles (Sternberg, 1985). For example, consider the following analogy: LAWYER:CLIENT:DOCTOR: ___ This problem should be read as ‘LAWYER is to CLIENT as DOCTOR is to ___’. Apprehension of experience refers to people’s ability to perceive and understand what they experience; thus, reading and understanding each of the words in the analogy requires apprehension of experience. Eduction (not ‘education’) is the process of drawing or bringing out, that is, making sense of, given facts. In this case, eduction of relations refers to the ability to perceive the relation between lawyer and client; namely, that the lawyer works for and is paid by the client. Eduction of correlates refers to the ability to apply a rule inferred from one case to a similar case. Thus, the person whom a doctor works for and is ultimately paid by is obviously a patient. Because analogy problems require all three of Spearman’s principles of cognition, he advocated their use in intelligence testing. Empirical evidence for Spearman’s two-factor theory comes from correlations among various tests of intellectual abilities. The governing logic is as follows. If we administer 10 different tests of intellectual abilities to a group of people and each test measures a separate, independent ability, the scores these people make on any one test will be unrelated to their scores on any other; the correlations among the tests will be approximately zero. However, if the tests measure abilities that are simply different manifestations of a single trait, the scores will be related; the intercorrelations will be close to 1. In fact, the intercorrelations among a group of tests of intellectual abilities are neither zero nor 1. Instead, most of these tests are at least moderately correlated, so that a person who scores well on a vocabulary test also tends to score better than average on other tests, such as arithmetic or spatial reasoning. The correlations among various tests of intellectual ability usually range from 0.3 to 0.7, which means that they have between 9 per cent and 49 per cent of their variability in common (Ozer, 1985). Spearman concluded that a general factor (g) accounted for the moderate correlations among different tests of ability. Thus, a person’s score on a particular test depends on two things: the person’s specific ability (s) on the test (such as spatial reasoning) and their level of the g factor, or general reasoning ability

Answer 14

Factor analysis is a statistical procedure developed by Spearman and Pearson that permits investigators to identify common factors among groups of tests. You might remember this from Chapter 2 when we discussed it in the context of items on questionnaires. It is a form of data reduction in the sense that a large number of data can be reduced and explained by reference to two or three factors (Kline, 1993) and is widely used in research that administers questionnaire measures. In the case of intelligence tests, these common factors would be abilities that affect people’s performance on more than one test. If a group of people take several different tests of intellectual ability and each person’s scores on several of these tests correlate well with one another, the tests may (at least partly) be measuring the same factor. A factor analysis determines which sets of tests form groups. For example, Birren and Morrison (1961) administered the Wechsler Adult Intelligence Scale (WAIS, an intelligence test described in the next section) to 933 people. This test consists of 11 different subtests. Birren and Morrison calculated the correlations between subtests and then subjected these correlations to a factor analysis. Table 11.1 shows the results of the analysis. The factor analysis revealed three factors, labelled A, B and C. The numbers in the three columns in the table are called factor loadings; they are somewhat like correlation coefficients in that they express the degree to which a particular test is related to a particular factor. For the various subtests on factor A, the largest factor loading is for vocabulary, followed by information, comprehension and similarities. In the middle range are picture completion, arithmetic, picture arrangement and digit symbol. Digit span, object assembly and block design are the smallest. Verbal subtests make the most important contribution to factor A, so we might be tempted to call this factor verbal ability. But almost all tests make at least a moderate contribution, so perhaps this factor may reflect general intelligence. Digit span has a heavy loading on factor B (0.84), and arithmetic and digit symbol have moderate loadings. Factor B, therefore, is related to maintaining information in short-term memory and manipulating numbers. Factor C appears to be determined mainly by block design, object assembly, picture completion and picture arrangement, and might, therefore, represent the factor, spatial ability. Although factor analysis can give hints about the nature of intelligence, it cannot provide definitive answers. The names given to factors are determined by the investigator and, although the names may appear to be quite appropriate, the process inevitably has a subjective element to it. There is also the danger of reification when conducting factor analysis. That is, the factors may wrongly be seen as concrete entities and not simply as labels used to describe a set of data as concisely and accurately as possible. Furthermore, factor analysis can never be more meaningful than the individual tests on which it is performed. To identify the relevant factors in human intelligence, one must include an extensive variety of tests in the factor analysis. The WAIS, for example, does not contain a test of musical ability. If it did, a factor analysis would undoubtedly yield an additional factor. Whether musical ability is a component of intelligence depends on how we decide to define intelligence; this question cannot be answered by a factor analysis. Other psychologists employed factor analysis to determine the nature of intelligence. Thurstone’s study (1938) of students’ performance on a battery of 56 tests extracted seven factors, which he labelled verbal comprehension, verbal fluency, number, spatial visualisation, memory, reasoning and perceptual speed. At first, Thurstone thought that his results contradicted Spearman’s hypothesised g factor. However, Eysenck suggested a few years later that a second factor analysis could be performed on Thurstone’s factors. If the analysis found one common factor, then Spearman’s g factor would receive support. In other words, if Thurstone’s seven factors themselves had a second-order factor in common, this factor might be conceived of as general intelligence. Cattell performed a second-order factor analysis and found not one but two major factors. Horn and Cattell (1966) called these factors fluid intelligence (gf) and crystallised intelligence (gc). Fluid intelligence is reflected by performance on relatively culture-free tasks, such as those that measure the ability to see relations among objects or the ability to see patterns in a repeating series of items. Crystallised intelligence is defined by tasks that require people to have already acquired information, such as vocabulary and semantic information, and is therefore more culture bound. Cattell regards fluid intelligence as closely related to a person’s native capacity for intellectual performance; in other words, it represents a potential ability to learn and solve problems. He regarded crystallised intelligence as what a person has accomplished using their fluid intelligence, that is, what they have learned. Horn (1978) disagreed with Cattell, citing evidence suggesting that both factors are learned but are also based on heredity. He says that gf is based on casual learning and gc is based on cultural, school-type learning. Figure 11.1 shows examples from four of the subtests that load heavily on fluid intelligence. Tests that load heavily on the crystallised intelligence factor include word analogies and tests of vocabulary, general information and use of language. According to Cattell, gc depends on gf . Fluid intelligence supplies the native ability, whereas experience with language and exposure to books, school and other learning opportunities develop crystallised intelligence. If two people have the same experiences, the one with the greater fluid intelligence will develop the greater crystallised intelligence. However, a person with a high fluid intelligence exposed to an intellectually impoverished environment will develop a poor or mediocre crystallised intelligence. Table 11.2 presents a summary of tests that load ongf and gc Although no two investigators may agree about the nature of intelligence, most believe that a small number of common factors account for at least part of a person’s performance on intellectual tasks.

Answer 15

Sternberg (1985) has devised a theory of intelligence that derives from the information-processing approach used by many cognitive psychologists. Sternberg’s theory has three parts; he calls it a triarchic theory (meaning ‘ruled by three’). The three parts of the theory deal with three aspects of intelligence: componential intelligence, experiential intelligence and contextual intelligence. Taken together, these three components go beyond the abilities measured by most common tests of intelligence. They include practical aspects of behaviour that enable a person to adapt successfully to their environment. Componential intelligence consists of the mental mechanisms people use to plan and execute tasks. The components revealed by the factor analyses of verbal ability and deductive reasoning are facets of componential intelligence. Sternberg suggests that the components of intelligence serve three functions. Meta-components (transcending components) are the processes by which people decide the nature of an intellectual problem, select a strategy for solving it and allocate their resources. For example, good readers vary the amount of time they spend on a passage according to how much information they need to extract from it (Sternberg, 1985). This decision is controlled by a meta-component of intelligence. Performance components are the processes used to perform the task, for example, word recognition and working memory. Knowledge acquisition components are those the person uses to gain new knowledge by sifting out relevant information and integrating it with what they already know. The second part of Sternberg’s theory deals with experiential intelligence. Experiential intelligence is the ability to deal effectively with novel situations and to automatically solve problems that have been previously encountered. According to Sternberg’s theory, a person with good experiential intelligence can deal more effectively with novel situations than a person with poor experiential intelligence. The person is better able to analyse the situation and to bring mental resources to bear on the problem, even if they have never encountered one like it before. After encountering a particular type of problem several times, the person with good experiential intelligence is also able to ‘automate’ the procedure so that similar problems can be solved without much thought, freeing mental resources for other work. A person who has to reason out the solution to a problem every time it occurs will be left behind by people who can give the answer quickly and automatically. Sternberg suggests that this distinction is closely related to the distinction between fluid and crystallised intelligence. According to Sternberg, tasks that use fluid intelligence are those that demand novel approaches, whereas tasks that use crystallised intelligence are those that demand mental processes that have become automatic The third part of Sternberg’s theory deals with contextual intelligence, that is, intelligence reflecting the behaviours that were subject to natural selection in our evolutionary history. Contextual intelligence takes three forms: adaptation, selection and shaping. The first form, adaptation, consists of fitting oneself into one’s environment by developing useful skills and behaviours. In different cultures, adaptation will take different forms. For example, knowing how to distinguish between poisonous and edible plants is an important skill for a member of a hunter–gatherer tribe. Knowing how to present oneself in a job interview is an important skill for a member of an industrialised society. The second form of contextual intelligence, selection, refers to the ability to find one’s own niche in the environment. That is, individuals will decide on careers or activities which they both enjoy doing and do well. The third form of contextual intelligence is shaping. Adapting to the environment or selecting a new one may not always be possible or profitable. In such cases, intelligent behaviour consists of shaping the environment itself. For example, a person whose talents are not appreciated by their employer may decide to start their own business

Answer 16

Gardner’s theory of intelligence is based on a neuropsychological analysis of human abilities (Gardner, 1983). It argues that intelligence falls into seven categories: linguistic intelligence, musical intelligence, logical/mathematical intelligence, spatial intelligence, bodily/kinesthetic intelligence and two types of personal intelligence. Bodily/kinesthetic intelligence includes the types of skill that athletes, typists, dancers or mime artists exhibit. Personal intelligence includes awareness of one’s own feelings (intrapersonal intelligence) and the ability to notice individual differences in other people and to respond appropriately to them, in other words, to be socially aware (interpersonal intelligence).

Answer 17

A recent theory which has attempted to explain why people who perform well on some tests of cognitive ability also perform well on others suggests that cognitive tests depend on executive processes such as working memory that are domain-general, and other processes that are domain-specific (Kovacs and Conway, 2016; 2019). As executive processes are required across a variety of different cognitive tasks, they are recruited more often than domain-specific ones. This is called process overlap theory – because these processes overlap – and is an attempt at explaining the positive manifold discovered by Spearman (1904), one of the most consistently replicated findings in psychology. The positive manifold refers to the phenomenon whereby variables are positively correlated. Spearman discovered that scores on some cognitive tests (e.g. arithmetic), as you've seen, correlated with others (e.g. mental rotation) and this correlation was explained with reference to a general intelligence factor: a latent variable that explained the underlying correlation between different types of abilities. It explains around 40 per cent of the variance in performance of a wide ranges of cognitive tests (Deary et al, 2010; although this does leave 60 per cent unaccounted for by general intelligence (Ackerman, 2016)). Except, of course, that some tests correlate more strongly than others; vocabulary and comprehension scores correlate well with each other but these do not correlate with mental rotation scores. As Kovacs and Conway write, ‘The problem with g is simply that still to this day there is no satisfactory consensus about how to interpret it: If there is a causal factor behind g, it has not been identified yet. . . there is substantial confusion about what kind of thing g, or indeed what any latent variable, is in the first place’ (p152)

Answer 18

A different type of intelligence, one not based on any particular cognitive ability, was proposed by Goleman (1995, 1998). This type of intelligence refers to the social and emotional components of interactions with others: the more socially sensitive and emotionally sensitive you were to the needs and behaviours of others, the more successful your interaction would be. Goleman referred to this as emotional intelligence but there is little evidence that this is a separate, valid and reliable type of intelligence (Sjoberg, 2001).

Answer 19

water level test - If you are a man, you probably drew the line horizontally across the glass; if you are a woman, you probably drew the line parallel to the direction in which the water glass is tipped. The correct line would be the horizontal one. This phenomenon illustrates one of the most consistent sex differences in cognitive ability. The task is Piaget’s Water Level Test, and men tend to be better at it than women (Halpern, 1992; Rilea, 2008). Why should males tend to be better at this task? One explanation for the Water Level Test result is that men and boys are intrinsically superior at tests of spatial ability than are women and girls (we will come on to reasons why this should be a little later). Spatial ability refers to ‘skills in representing, transforming, generating and recalling symbolic, non-linguistic information’ (Linn and Petersen, 1985). The test of spatial ability which shows the most consistent and reliable sex difference is mental rotation (Masters and Sanders, 1993). In this task, individuals are presented with three sets of cubes and have to match the target set with one of the other two. The task is not straightforward because the cubes have to be mentally rotated before a match can be made (see Chapter 8) A different version of the standard mental rotation task was used in a study by Hirnstein et al (2009). The researchers presented the well-known Vandenberg and Kuse mental rotation test in its original and in modified form. In the original, the participant is presented with one target and four sample sets of cubes and is asked to identify two samples that are identical, when rotated, to the target set. In the modified version, the participant has to compare each sample against the target, a modification that prohibits a leaping strategy, that is, moving onto the next trial as soon as they think they have identified two matching samples (thereby ignoring the remaining, uncompared samples). Hirnstein et al (2009) found that the performance of both men and women was poorer in the modified condition, but men’s performance was more adversely affected than that of women. However, the better performance in men remained. However, object location memory is thought to be better in women, a finding that has been replicated internationally in 35 out of 40 countries (Silverman et al, 2007). Honda and Nihei (2009) asked men and women to study a variety of objects and to recall the location of the objects either three minutes after presentation or one week later. Women were better than men after three minutes but not after a week. They were also better at locating objects whose locations had been swapped but, again, only in the three-minute condition. In other types of visuospatial test, women tend to outperform men. For example, women have been found to be consistently superior to men at tests involving visual recognition. It has been suggested that this may be so because of women’s superior linguistic ability. Women, for example, are better than men at tests involving verbal fluency, such as naming as many objects as possible beginning with a specific letter (Halpern, 1997), and boys tend to be diagnosed with reading and speech disorders more commonly than are girls (Flynn and Rahbar, 1994). The stimuli used can also be important. Rilea (2008) found that men were better at the Water Level Task but showed no advantage on mental rotation or paper folding. However, men were better at rotating polygons than stick figures (and they showed a right hemisphere advantage for the task; there was no right hemisphere advantage shown by women for either stimulus type). Contreras et al (2007) suggest a stimulus-specific complicating factor: the tests of spatial ability used are static, not dynamic. Therefore, they set up a study in which men and women performed a dynamic test of spatial ability: guiding two dots towards a destination in the shortest time. They found that even when performance factors were taken into account, men outperformed women (i.e. differences in performance style did not explain the sex difference). On average, men took longer to decide on the first move of the dots, then guided the dots more quickly. There are also interactions between handedness and the visuospatial rotation of specific stimuli. Right-handers, for example, are faster at mentally rotating hands than are left-handers (although they are not more accurate) (Jones et al, 2021). Right-handers are also quicker at identifying images of right hands.

Answer 20

Theories of sex differences in cognitive ability fall into four general categories: evolutionary, psychosocial, biological and cognitive

Answer 21

The evolutionary point of view suggests that spatial superiority in men is a throwback to the evolution of men and women as hunters and gatherers (Eals and Silverman, 1994). This theory suggests that because men originally roamed and hunted (activities which rely on the manipulation of visuospatial features in the environment), and because women stayed ‘at home’ and gathered, it is not surprising that men are spatially superior. The greater visual recognition performance seen in women is meant to reflect women’s evolutionary role as foragers (Tooby and DeVore, 1987). According to evolutionary psychologists, one reason why men are better at spatial cognition than women is that men were the hunters who ranged far and wide for their prey and would, therefore, need to develop a well-tuned set of navigational skills. Women, the child-bearers and rearers, stayed at home and foraged. Some argue that women’s ranging was limited to picking plants; men would hunt for game. According to Ecuyer-Dab and Robert (2004), however, this dichotomy suggests that rather than showing a superior spatial advantage by men over women it shows how context can affect the way in which each sex expresses its specific spatial skills: spatial cognition in men would be used to navigate the environment for a mate and food whereas women’s spatial cognition developed to deal with the immediate environment because they were more concerned with the survival of their offspring in the home. They, therefore, had no need to develop the navigational spatial skills that men did. In short, men developed and evolved large-scale navigation mechanisms and women evolved small-scale ones.

Answer 22

Helena Cronin, the evolutionary psychologist, has pithily described the male species as having more Nobels but also more dumbbells. That is, men are more likely to show extreme performances on measures of intelligence and cognitive ability whereas women may show a less extreme pattern. There is some evidence of this. An extensive study of 320,000 11–12-year-olds in the UK, found few strong differences between girls and boys on the Cognitive Abilities Test but boys were more strongly represented at the top and bottom of the distribution for the non-verbal measure scores and at the lower end of the verbal measure distribution (Strand et al, 2006). Lohman and Lakin (2009) administered the same test in three different versions to over 318,000 children from grades 3 to 11 in North America. They found almost identical results to those reported by the UK study, suggesting that the findings cannot be attributable to nation, age, education system or type of test. Instead, the results suggest a more universal finding. Ecuyer-Dab and Robert cite evidence that women, for example, were more likely than men to use landmarks when giving map directions. Men are more likely to provide more detail on direction and distance – although women are capable of doing this, they simply do not use these references as their primary source of information. One objection to the theory that males are intrinsically superior to females on tests of mental rotation is that the results may be attributable to other causes. For example, because the task is timed, it has been argued that this is detrimental to women, who are more cautious when making decisions about rotation (Goldstein et al, 1990). tTo test this hypothesis, Masters (1998) allowed male and female undergraduates either a short or unlimited time to perform a mental rotation task. She also used three different scoring procedures because previous studies had been criticised for basing their findings on using correct answers only (without looking at the number of incorrect responses too). Masters found that regardless of scoring procedure or time limit, men performed better than women. (Interestingly, however, some sex differences, such as female self-reported confusion over left and right, may be attributable to women rating themselves more critically than do men (Jordan et al, 2006).) However, evolutionary theories such as these are so broad as to be untestable (see Chapter 3). As Halpern (1997) also notes, you can explain almost any finding by indicating how it would be advantageous to hunters and gatherers.

Answer 23

Psychosocial theories suggest that sex differences are learned through experience or imitation. Children, it is argued, fulfil sex-role stereotypes: boys are encouraged to play with toys which involve visuospatial manipulation; girls are not (we will come back to this in Chapter 12). It has also been suggested that boys and girls receive different models, rewards and punishment. One researcher has suggested that peer interaction is more likely to lead to stereotypical sex-role behaviour than is parent–child interaction, although this idea is controversial (Harris, 1995). Halpern (1992) cites fairly strong evidence against a psychosocial explanation for sex differences in cognitive ability. She noted that among individuals with high reasoning ability, right-handed men outperformed left-handed men on tests of spatial ability but were poorer than left-handed men at verbal tasks. Conversely, left-handed females were better at spatial tasks than were right-handers, but the opposite pattern applied to verbal tasks. Any theory of psychosocial influence would have difficulty in explaining these findings: why should right- and left-handed boys and girls be socialised differently? It would also have difficulty in explaining why boys are more likely than girls to suffer from stuttering and reading disorders

Answer 24

Biological theories suggest that sex differences in cognitive ability may be due to biological factors such as hormonal regulation and brain organisation. Keller and Menon (2009) examined brain activation and structure in 25 men and women who performed various mathematical operations including subtraction and addition. While men and women were equally accurate and equally as fast at the tests, their brain activation during processing differed. There was greater activity in the right dorsal and ventral streams in men and in an area of the right parietal lobe known to be important for calculating arithmetical problems. In terms of structure, women had greater neuronal density in these areas compared with men. The authors suggest that the differences reflect women’s more efficient use of neural resources. Apart from neuroanatomical differences, there may also be differences in the amount of, or sensitivity to, hormones expressed by women and men (Collaer and Hines, 1995). However, other studies have also found no relationship between hormone level and spatial ability (Liben et al, 2002). This may not mean that steroids are not involved. ‘It may be’, as the authors suggest, ‘that such effects do occur but only under some as yet unidentified additional setting conditions (be they biological or experiential)’.

Answer 25

Assessment of intellectual ability, or intelligence testing, is a controversial topic because of its importance in modern society. Unless people have special skills that suit them for a career in sports or entertainment, their economic success may depend heavily on formal education. Many employers use specialised aptitude tests to help them select among job candidates. Test scores correlate with school and university grades, the number of years in education and adult occupational status (Nisbett et al, 2012) but whether they make you a star appointment is unclear. There are hundreds of tests of specific abilities, such as manual dexterity, spatial reasoning, vocabulary, mathematical aptitude, musical ability, creativity and memory. All these tests vary widely in reliability, validity and ease of administration. And as you saw in Chapter 4, some of these tests show evidence of cultural bias

Answer 26

Intelligence testing has a long and chequered history. As early as 2200 BC, Chinese administrators tested civil servants (mandarins) periodically to be sure that their abilities qualified them for their job. In Western cultures, differences in social class were far more important than individual differences in ability until the Renaissance, when the modern concept of individualism came into being The term ‘intelligence’ is an old one, deriving from the Latin intellectus (meaning ‘perception’ or ‘comprehension’). However, its use in the English language dates only from the late nineteenth century when it was revived by the philosopher Herbert Spencer (1820–1903) and by the biologist/statistician Sir Francis Galton (1822–1911). Galton was the most important early investigator of individual differences in ability. He was strongly influenced by his cousin Charles Darwin, who stressed the importance of inherited differences in physical and behavioural traits related to a species’ survival. Galton observed that there were family differences in ability and concluded that intellectual abilities were heritable. Having noted that people with low ability were poor at making sensory discriminations, he decided that tests involving such discriminations would provide valid measures of intelligence. In 1884, Francis Galton established the Anthropometric Laboratory (meaning ‘human-measuring’) at the International Health Exhibition in London. His exhibit was so popular that afterwards his laboratory became part of the South Kensington Museum. He tested over 9,000 people on 17 variables, including height and weight, muscular strength and the ability to perform sensory discriminations. One task involved detecting small differences in the weights of objects of the same size and shape. Galton made some important contributions to science and mathematics. His systematic evaluation of various large numbers of people and the methods of population statistics he developed served as models for the statistical tests now used in all branches of science. His observation that the distribution of most human traits closely resembles the normal curve (developed by the Belgian statistician Lambert Quetelet, 1796–1874) is the foundation for many modern tests of statistical significance Galton also outlined the logic of a measure he called correlation: the degree to which variability in one measure is related to variability in another. From this analysis, the British mathematician Karl Pearson (1857–1936) derived the correlation coefficient (r) used today to assess the degree of statistical relation between variables. In addition, Galton developed the logic of twin studies and adoptive parent studies to assess the heritability of a human trait. Galton, however, has a troubling relationship with psychology and with humanity generally. Although a polymath, one of his creations is notoriously better-known than any of his others. He coined the term “eugenics”, for example, and was a President of the British Eugenics Society. He was not alone in his beliefs, Maynard Keynes and Winston Churchill held similar views (Langkjaer-Bain, 2019) and a belief in eugenics and innate abilities and superiority were commonly held at the time

Answer 27

Alfred Binet (1857–1911), a French psychologist, and a colleague (Binet and Henri, 1896) suggested that a group of simple sensory tests could not adequately determine a person’s intelligence. They recommended measuring a variety of psychological abilities (such as imagery, attention, comprehension, imagination, judgements of visual space, and memory for various stimuli) that appeared to be more representative of the traits that distinguished people of high and low intelligence. To identify children who were unable to profit from normal classroom instruction and needed special attention, Binet and Théodore Simon assembled a collection of tests, many of which had been developed by other investigators, and published the Binet–Simon Scale in 1905. The tests were arranged in order of difficulty, and the researchers’ obtained norms for each test. Norms are data concerning comparison groups that permit the score of an individual to be assessed relative to his or her peers. In this case, the norms consisted of distributions of scores obtained from children of various ages. Binet and Simon also provided a detailed description of the testing procedure, which was essential for obtaining reliable scores. Without a standardised procedure for administering a test, different testers can obtain different scores from the same child. Binet revised the 1905 test to assess the intellectual abilities of both normal children and those with learning problems. The revised versions provided a procedure for estimating a child’s mental age, that is, the level of intellectual development that could be expected for an average child of a particular age. For example, if an 8-year-old child scores as well as average 10-year-old children, their mental age is 10 years. Binet did not develop the concept of IQ. Nor did he believe that the mental age derived from the test scores expressed a simple trait called ‘intelligence’. Instead, he conceived of the overall score as the average of several different abilities.

Answer 28

Lewis Terman of Stanford University translated and revised the Binet–Simon Scale in the USA. The revised group of tests, published in 1916, became known as the Stanford–Binet Scale. Revisions by Terman and Maud Merrill were published in 1937 and 1960. In 1985, an entirely new version was published. The Stanford–Binet Scale consists of various tasks grouped according to mental age. Simple tests include identifying parts of the body and remembering which of three small cardboard boxes contains a marble. Intermediate tests include tracing a simple maze with a pencil and repeating five digits orally. Advanced tests include explaining the difference between two abstract words that are close in meaning (such as fame and notoriety) and completing complex sentences. The 1916 Stanford–Binet Scale contained a formula for computing the intelligence quotient (IQ), a measure devised by Stern (1914). The IQ represents the idea that if test scores indicate that a child’s mental age is equal to their chronological age (i.e. calendar age), the child’s intelligence is average; if the child’s mental age is above or below their chronological age, the child is more or less intelligent than average. This relation is expressed as the quotient of mental age (MA) and chronological age (CA). The result is called the ratio IQ. The quotient is multiplied by 100 to eliminate fractions. For example, if a child’s mental age is 10 and the child’s chronological age is 8, then their IQ is (10/8) × 100 = 125. The 1960 version of the Stanford–Binet Scale replaced the ratio IQ with the deviation IQ. Instead of using the ratio of mental age to chronological age, the deviation IQ compares a child’s score with those received by other children of the same chronological age (the deviation IQ was invented by David Wechsler, whose work is described in the next section). Suppose that a child’s score is one standard deviation above the mean for their age. The standard deviation of the ratio IQ scores is 16 points, and the score assigned to the average IQ is 100 points. If a child’s score is one standard deviation above the mean for their age, the child’s deviation IQ score is 100 + 16 (the standard deviation) = 116. A child who scores one standard deviation below the mean receives a deviation IQ of 84 (100 – 16)

Answer 29

When David Wechsler was chief psychologist at New York City’s Bellevue Psychiatric Hospital he developed several popular tests of intelligence. The Wechsler–Bellevue Scale, published in 1939, was revised in 1942 for use in the armed forces and was superseded in 1955 by the Wechsler Adult Intelligence Scale (WAIS). This test was revised again in 1981 (the WAIS-R), 1997 (the WAIS-III) and 2008 (WAIS-IV). The Wechsler Intelligence Scale for Children (WISC) was first published in 1949; the fifth edition was published in 2016. Various versions of the WAIS-R have been devised for use with various populations (such as Irish, Scottish, Welsh, and so on). Previous versions of the scale provided a measure called ‘full-scale IQ’ which comprised scores from two separate subscales – performance IQ and verbal IQ. The current version, the WAIS-IV, however, has dispensed with the two subscales and now provides a total full-scale IQ score. It is a large collection of individual tests (the test is called a ‘battery’) currently validated on 2,200 individuals between 16 and 90 years of age. The tests which form the WAIS-IV can be seen in Figure 11.6a. You can see that full-scale IQ is made up of scores from four separate subscales all of which have core components or subtests. It is these core components (10 tests) which contribute to full-scale IQ. Figure 11.6b and c shows you examples of two of the new tests in the WAIS-IV that did not feature in WAIS-III. The WAIS is the most widely administered adult intelligence test

Answer 30

The adequacy of a measure is represented by its reliability and validity (terms described in Chapter 2). In the case of intelligence testing, reliability is assessed by the correlation between the scores that people receive on the same measurement on two different occasions; perfect reliability is 1. High reliability is achieved by means of standardised test administration and objective scoring: all participants are exposed to the same situation during testing, and all score responses in the same way. The acceptable reliability of a modern test of intellectual ability should be at least 0.85. Validity is the correlation between test scores and the criterion, that is, an independent measure of the variable that is being assessed. However, most tests of intelligence correlate reasonably well with such measures as success in school (between 0.40 and 0.75). Thus, because intellectual ability plays at least some role in academic success, IQ appears to have some validity

Answer 31

One of the most curious phenomena of intelligence measurement is that people appear to be getting significantly more intelligent or, more accurately, their IQ scores are increasing, at least in some countries. This is called the Flynn effect after the psychologist James Flynn who noticed that the average level of intelligence in the US had risen since the beginning of the twentieth century: a 20-point acceleration in IQ in 60 years (Flynn, 1984, 1987). Although this term has stuck, Flynn was not the first to observe generational increases in IQ/intelligence test performance; around 24 studies had been published before Flynn’s widely cited study, and the first study showing the pattern was reported in 1936 (Lynn, 2013). That IQ is increasing has also been the subject of some scepticism (see below). Why would IQ increase in this way, even though the increase appears to be attenuating? The most obvious explanation is improved schooling and education, especially in primary or elementary school where the focus has shifted towards teaching cognitive skills that reflect fluid intelligence (Blair et al, 2005a). Some suggest that this improvement may have halted in economically developed countries in the recent two decades. Lynn (2009) exploited the fact that measures of intelligent behaviour have been standardised in the past few years (standardisation is the process whereby a test’s validity is tested to ensure that it is a measure of what it claims to measure for a given population). One, the well-known Raven’s Coloured Progressive Matrices, was standardised in the UK in 2007 and 2008. In the task, participants have to select from several options the one which would complete a sequence with the final part missing. Over the period 1982–2007, there was a difference between children between 4 and 11 years old of 8 IQ points. Children between 7 and 15 completing the Standard Progressive Matrices showed an 8.2 IQ point difference between 1979 and 2008 The Flynn effect appears to be fairly strong in countries such as Austria, France, Germany, Israel, Japan, Kenya, the Netherlands, Spain (and, more recently, the Czech Republic; Laciga and Cigler, 2016), but the effect is weaker in countries such as Australia, Brazil, Ireland, the UK, the US and New Zealand (Pietschnig and Voracek, 2015). Also, there are some contradictory findings emerging (see below). Some countries (Sweden, Norway) appear to show no current increases in IQ and some others (Denmark, Finland) appear to show a reversal, with IQ points dropping. A meta-analysis of data from 271 samples (4 million people), from 31 countries, collected between 1909 and 2013 found that when the Flynn effect did occur, it was very IQ-domain-specific; that is, it was stronger for fluid intelligence (where there have been substantial gains) and weaker for crystallised intelligence (Pietschnig and Voracek, 2015). It was also stronger in adults, than in children, and has decreased in recent decades. They found that gains in crystallised intelligence had stopped in the period 1987–2011. Fluid intelligence had gained by 0.43 IQ point every year until 1985 and by 0.22 points a year up until 2013. Dutton & Lynn (2015) found that IQ/intelligence in France had declined. This was based on a drop in scores from the third iteration of the WAIS (WAIS-III) to the fourth (WAIS-IV). This caused quite a reaction. ‘The IQ of the French in freefall’, reported one media outlet. Gonthier et al (2021), however, suggested that the decline was not a genuine drop and was the result of biased items and limited to subscales of the WAIS (which skewed overall intelligence scores). There is also the issue that the 2015 study was based on a sample of 79 people They found that only subscales which involved cultural knowledge showed a decline over time. The Verbal Comprehension Index, for example, included items testing declarative memory for material specific to a culture. The main difference between the scores from WAIS-III and WAIS-IV was attributable to the Vocabulary, Comprehension and Information subscales of the WAIS. There were also large differences in the content of the tests used in the 3rd and 4th editions. In the French version of WAIS-IV, 11 of 18 comprehension items were changed, 29/33 of the vocabulary items and 21/28 of the Information items. Meanwhile, in the US, something seems to be afoot with vocabulary. Twenge et al (2019) analysed data from 29, 912 individuals who were part of the General Social Survey of US adults, a representative survey of US residents which has been conducted since 1974. When controlling for levels of education, Twenge et al found that vocabulary scores had declined from the mid-1970s to the 2010s. They reported that the largest drop was in those with degrees, suggesting that education itself may not be particularly well-correlated with education. The result seems like an anti-Flynn pattern: a decline, rather than an improvement. Twenge et al note that Flynn improvements are typically reported for non-verbal abilities and that the Flynn effect may no longer be consistent; Germany, for example, appears to be showing a reverse Flynn effect (Pietschnig & Gittler, 2015)

Answer 32

Does spending more time – and more time successfully – in education result in improved IQ? Ritchie and Tucker-Drob (2018) suggests it does. They conducted a meta-analysis on what they described as three categories of quasi-experimental studies of education and intelligence. From 42 data sets involving 600,000 participants, they found that education was associated with improved greater intelligence to the point where they estimate that a person might add 1-5 IQ points for every year of education. They asked a number of questions of their data such as: (1) whether the effects are additive, that is, does increasing education ‘add’ to intelligence, or is this implausible because a plateau at some point would have to be reached; (2) whether there is a Matthew effect, that is, brighter children benefit most from education and so become brighter or an improvement effect in that children with lower initial ability improve most from continuing education. They also considered performance on which cognitive tests are most improved; they noted that the studies included in the meta-analysis used a variety of tests and did not measure latent g. And there remains the (additional) unanswered question: what is the mechanism underlying this improvement? There is a significant association between national IQ and educational achievement. Lynn and Mikk (2007) correlated published national IQ scores in 2002 and 2006 with the educational attainment scores of children in 25 (10-year-olds) and 46 (14-year-olds) countries. Maths and science achievement correlated significantly with national IQ. IQ was also associated with per capita income: the greater the IQ, the greater the income. Some studies, with small samples, have also found that adolescents who come from wealthier backgrounds have greater grey matter volume than those from poorer backgrounds but greater or lesser white matter (Mackey et al, 2015). Cortical thickness was greater in all lobes in the wealthier group and thickness in the temporal and occipital lobes was positively correlated with a standardised test of cognitive ability (the Massachusetts Comprehensive Assessment System)

Answer 33

Some abilities – intellectual, athletic, musical and artistic – appear to run in families. Why? Are the similarities owing to heredity, or are they solely the result of a common environment, which includes similar educational opportunities and exposure to people having similar kinds of interests? According to Sternberg and Grigorenko (1997), we know three facts about the roles of heredity and environment in intelligence: (1) both contribute to intelligence; (2) they interact in various ways; and (3) poor and enriched environments influence the development of intellectual ability regardless of heredity. What these facts illustrate is that the typical nature–nurture debate in intelligence is no longer valid. The nature–nurture argument suggests that, in its most stark form, behaviour or function is determined solely by the environment or solely by genetics/heredity. Psychologists have discovered that this argument is too simplistic. In fact, it is inaccurate. Almost all psychologists agree that intelligence has a hereditary (as well as environmental) component. The debate now focuses on the degree to which each contributes to intelligence and the ways in which they interact to influence intellectual development.

Answer 34

When we ask how much influence heredity has on a given trait, we are usually asking what the heritability of the trait is, as you saw in the genetics section of Chapter 3. Heritability is a statistical measure that expresses the proportion of the observed variability in a trait that is a direct result of genetic variability (GV). The value of this measure can vary from 0 to 1. The heritability of many physical traits in most cultures is very high; for example, eye colour is affected almost entirely by hereditary factors and little, if at all, by the environment. Thus, the heritability of eye colour is close to 1. Heritability is a concept that many people misunderstand. It does not describe the extent to which the inherited genes are responsible for producing a particular trait; it measures the relative contributions of differences in genes and differences in environmental factors to the overall observed variability of the trait in a particular population. As with hair colour, we infer the heritability of a person’s intelligence from their observed performance. Thus, looking at a person’s IQ score is equivalent to looking at the colour of a person’s hair. By measuring the correlation between IQ score and various genetic and environmental factors, we can arrive at an estimate of heritability. Clearly, even if hereditary factors do influence intelligence, the heritability of this trait must be considerably less than 1 because so many environmental factors also influence intelligence. The branch of psychology called behaviour genetics (see Chapters 1 and 3), predicts the degree of parental influence via genetic and environmental transmission on the development of the child’s intellectual development. The proportion of the variance associated with genetic differences among individuals is called h; the remaining variation which is associated with environmental influences is referred to as 1 – h (Neisser et al, 1996a). The features which families share and have in common (such as choice of home) is sometimes referred to as c. Factor h can be subdivided into two types: additive h, which refers to the amount of hereditary variance that is passed from parent to child, and non-additive h, which refers to new, unique genetic expression in each generation. As children grow older, h increases and c decreases (McGue et al, 1993). In childhood, the contribution of h and c to intelligence is similar; by adolescence, h predicts about three-quarters of intellectual ability. The heritability of a trait depends on the amount of variability of genetic factors in a given population. If there is little GV, genetic factors will appear to be unimportant. Because the ancestors of people living in developed Western nations came from all over the world, GV is likely to be much higher there than in an isolated tribe of people in a remote part of the world. Therefore, if a person’s IQ score is at all affected by genetic factors, the measured heritability of IQ will be higher in, say, Western European culture than in an isolated tribe. The relative importance of environmental factors in intelligence depends on the amount of environmental variability (EV) that occurs in the population. If EV is low, then environmental factors will appear to be unimportant. In a society with a low variability in environmental factors relevant to intellectual development – one in which all children are raised in the same way by equally skilled and conscientious carers, all schools are equally good, all teachers have equally effective personalities and teaching skills, and no one is discriminated against – the effects of EV would be small and those of GV would be large. In contrast, in a society in which only a few privileged people receive a good education, environmental factors would be responsible for much of the variability in intelligence: the effects of EV would be large relative to those of GV.

Answer 35

Biological and environmental factors can affect intellectual abilities prenatally and post-natally. Newborn infants cannot be said to possess any substantial intellectual abilities; rather, they are capable of developing these abilities during their lives. Therefore, prenatal influences can be said to affect a child’s potential intelligence by affecting the development of the brain. Factors that impair brain development will necessarily also impair the child’s potential intelligence. As the axons of developing neurons grow, they thread their way through a tangle of other growing cells, responding to physical and chemical signals along the way. During this stage of prenatal development, differentiating cells can be misguided by false signals. For example, if a woman contracts German measles during early pregnancy, toxic chemicals produced by the virus may adversely affect the development of the foetus. Sometimes, these chemicals can misdirect the interconnections of brain cells and produce intellectual difficulties. Thus, although development of a human organism is programmed genetically, environmental factors can affect development even before a person is born. Educational influences in the environment, including (but not limited to) schooling, significantly affect the development of cognitive ability. Nisbett et al (2012) note that children who miss a year of school demonstrate a drop in their IQ score, compared to attenders. A child who enters the 5th grade (years 10–11) a year earlier than a child of the same age in the 4th grade (years 9–10) has a verbal IQ that is five points higher at the end of the school year. In Norway, there are indications that adding two years beyond the 7th grade (years 12–13) affects (improves) IQ at age 19 (Brinch and Galloway, 2011). According to Nisbett (2009), lengthening the school day, decreasing class sizes and using interactive computer games were all found to lead to an increase in academic skill Around 40 per cent of variance in educational attainment is attributable to genetic influence.

Answer 36

Estimates of the degree to which heredity influences a person’s intellectual ability come from several sources. The two most powerful methods are comparisons between identical and fraternal twins and comparisons between adoptive and biological relatives (see Chapter 3)

Answer 37

Evidence indicates that the heritability of IQ is between 0.4 and 0.8. That is, highly heritable (Nisbett et al, 2012). One reason for this may be that children select experiences that reflect their predispositions thus stimulating cognitive development. An alternative is that there are different genes that are activated at different stages in development (as they do for puberty) (Tucker-Drob et al, 2013). MZ twins are also more comparable in terms of brain volume and activation. Grey matter in the frontal, parietal and occipital cortex is virtually identical in volume in MZ twins. There is also considerable similarity in volume between DZ twins but more variation in the posterior occipital region and language-related frontal regions (Chiang et al, 2009; van Leeuwen et al, 2009). White matter volume is correlated in MZ twins but not in the frontal and parietal cortex in DZ twins. The structural pattern is mirrored in functional studies. For example, highly intelligent individuals given a simple or a moderately difficult problem to solve do so more quickly than less intelligent individuals and show less cortical activation, especially in the PFC (Neubauer and Fink, 2009). This probably reflects lack of effort required to solve simple puzzles. The contribution of h to intelligence appears to increase from 0.3 in early childhood (Cherny et al, 1994) to 0.8 in middle age (Finkel et al, 1995) and may extend to very old age (over 80 years of age) (Petrill et al, 1998). At least half the total variance in IQ scores is accounted for by genetic variance (Chipuer et al, 1990; Plomin et al, 1997). The fact that, by most estimates, genetic factors account for approximately 50 per cent of the variability in IQ scores means that the other half of the variability is accounted for by environmental factors. However, the contribution of the environment is less than 25 per cent. Some estimates, based on comparisons of parents and their offspring raised together or apart, suggest a value of only 4 per cent. Why are these figures so low? Plomin (1988) suggests that estimates of the importance of environmental factors tend to be low because the environment in a given family is not identical for all its members. Some environmental variables within a family are shared by all members of the family, such as the number of books the family has, the examples set by the parents, the places the family visits on holiday, the noisiness or quietness of the home, and so on. But not all of the environmental factors that affect a person’s development and behaviour are shared in this way. For example, no two children are treated identically, even by family members; differences in their appearances and personalities affect the way other people treat them. Different members of a family will probably have different friends and acquaintances, attend different classes in school and, in general, be exposed to different influences. And once people leave home, their environments become even more different. Plomin et al (2013), investigating the reasons for exceptional reading ability in the top 5 per cent of 10,000 12-year-old twins, found that genetic factors explained more than half the difference in performance between the exceptional and normal readers; growing up in the same family/attending the same school accounted for less than a fifth. Kovas et al (2013) found that numeracy and literacy were more heritable than was intelligence during early to late childhood (7–12 years). One reason for this is that schooling may reduce the variance attributable to environment, but as the children select and create their own environment later, cognitive ability correlates with heritability. Estimates of the contribution of EV to intelligence based on measurements made during childhood also tend to be higher than similar estimates based on measurements made during adulthood. The reason for this difference may be that, during childhood, family members share a more similar environment, whereas during adulthood their environments become less similar. As Plomin (1997) notes, studies of genetically unrelated children (of a mean age of under 10 years) adopted and raised in the same families, suggest that up to 30 per cent of the variability in IQ scores is due to common environmental factors. However, when the comparison is made among young adults, the figure drops to less than 3 per cent. Adopted children appear to become more like their birth parents but do not become more like their adoptive parents, findings that were also reported in a study of the development of antisocial behaviour in twins (Pike et al, 1996). Thus, once children leave home and are exposed to different environmental variables, the effect of a common family environment almost disappears. What is left, in the case of related individuals, is their common genetic heritage

Answer 38

Given that our DNA is what makes us what we are, the first and most obvious locus of any genetic cause would be our chromosomes (Plomin, 1997). The DNA contains sequences of information that are divided into sections by enzymes (called restriction enzymes). Sometimes these sequences are repeated. These enzymes act as markers that can be used to locate chromosomes and defects on chromosomes. Some genetic disorders of behaviour are single-gene disorders, that is, only one chromosome is affected. More complex behaviours, however, are likely to have multiple genetic loci called quantitative trait loci. The question, therefore, is whether intelligence is inherited through one gene or multiple genes. Recently, Genome-Wide Association Studies (GWAS; see Chapter 3) have been employed to try and clarify the role of genes in intelligence but, as Plomin et al (2013) note, if all the genes had been identified, there would not be any need for twin or adoption studies. The studies conclude that there are no genes which produce effects large enough to be meaningfully and consistently associated with intelligence and this suggests that the genes are likely to be many in number, each having small effect sizes. The first GWAS study, for example, found that the gene with the largest effect size accounted for less than 0.005 of the variance in cognitive ability. A study of 101,069 individuals and a replication sample of 25,490 identified three polymorphisms associated with educational attainment, that explained around 2 per cent of the variance (Rietveld et al, 2013). One meta-analysis of 16 studies that included over 9,000 participants, found that a polymorphism on one gene was associated with IQ. Another possible polymorphism associated with intelligence is Val66Met. Determining which genes, if any, are responsible for cognition is fraught with problems. Cognition is loosely defined, is made up of a variety of different components and is likely to be moderated not by a single gene but by a vast collection of genes. The polymorphisms that have been identified do not explain much of the variance in cognitive test scores (Davies et al, 2016). A study by Marioni et al (2018) examined this issue using a different approach; they used the epigenome to try and determine the genes involved in cognitive ability. The epigenome regulates genes by modifying DNA. The methylation of DNA, for example, involved adding a methyl group to a cytosine nucleotide placed next to a guanine nucleotide (Marioni et al, 2018). The sample was derived from 11 cohorts who had provided DNA and cognitive performance data from several international locations. Seven cognitive tests were included in the analysis, including several from the WAIS and Wechsler Memory Scale, the Mini Mental State Examination (a measure of general cognitive ability), verbal fluency tests, and various naming tests. he study found a significant association between global cognition scores and part of the gene labelled cg21450381 and between verbal fluency and one labelled cg12507869. These were located on Chromosomes 12 and 10

lecture 5 - personality and intelligence Flashcards

(62 cards)