Criticism of Intelligence Testing

11 Key excerpts on "Criticism of Intelligence Testing"

• 

  eBook - PDF

  Intelligence Testing and Minority Students

  Foundations, Performance Factors, and Assessment Issues

  • Richard R. Valencia, Lisa A. Suzuki(Authors)
  • 2000(Publication Date)
  • SAGE Publications, Inc

    (Publisher)

  Test Bias 115 with the psychometric paradigm itself, which views humans as fairly stable in cognitive functioning and measured intelligence as a culturally independent trait. By sharp con-trast, the cultural-psychology perspective asserts that the role of culture in intelligence and its measurement are critical to understand (see, e.g., Miller, 1997). Although it is quite difficult to operationalize and measure cultural loading in intel-ligence tests, there are objective and empirical methods to detect bias in such tests. When research on test bias began during the late 1970s, many scholars asserted that the con-struct of test bias can be explained in the context of validity theory and, as such, is an em-pirical, testable, quantifiable, and scientific matter. Novices to the subject of test bias typ-ically are surprised that the notion of bias is derived from mathematical statistics. In this field, the term bias refers to the systematic under- or overrepresentation of a population parameter by a statistic based on samples drawn from the population (Jensen, 1980, p. 375). In traditional psychometrics, however, bias takes on a related but distinct mean-ing. It typically is conceived as the systematic (not random) error of some true value of test scores that are connected to group membership (see, e.g., Jensen, 1980; Reynolds, 1982b). Note that group membership is the general referent. It can refer to test bias in the contexts of race/ethnicity, sex, social class, age, and so on. Test bias in the context of race/ethnicity often is referred to as cultural bias, the subject of this chapter. As we dis-cuss later, investigations of cultural bias in intelligence tests can, and sometimes do, em-ploy empirically defined and testable hypotheses and complex statistical analyses (see, e.g., Reynolds, 1982b). Another perspective on test bias was offered by Cole and Moss (1989).

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Debating Special Education

  • Michael Farrell(Author)
  • 2010(Publication Date)
  • Routledge

    (Publisher)

  Problems with assessment

  This chapter examines assessments, especially intelligence tests, used in special education. It first considers the nature of intelligence and the characteristics of typical tests of intelligence. The chapter next looks at the assessment of intelligence with regard to reading disorder and to cognitive impairment. It then reviews criticisms of intelligence and its assessment that:

  • the concept of intelligence and its testing are determinist; • intelligence is associated with separate lower status schooling for some pupils; • there are better alternatives.

  The nature of intelligence and characteristics of tests

  Intelligence has been said to concern various abilities. These are the ability to ‘understand complex ideas’, ‘adapt effectively to the environment’, ‘learn from experience’, ‘engage in various forms of reasoning’ and overcome obstacles by thinking about them (American Psychological Association Task Force, 1995). Individuals differ from one another in these abilities. If this conceptual outline is accepted, it follows that intelligent activity would involve related features such as seeing the essentials in a given situation and responding appropriately to them. There is debate about what responding appropriately might mean and what samples of such behaviour might be suitable to form part of any assessments of intelligence.

  When intelligence tests are administered, there tend to be variations in the scores of the subtests that contribute to the overall score. Nevertheless, the scores on these sub-tests tend to correlate positively, so that a person who scores highly on one sub test tends to score highly on others. Interpretations of the structure of intelligence have been informed by statistical methods used in interpreting data. The technique of factor analysis and different methods of ‘factoring’ the correlations between sub-tests of intelligence have led to various interpretations of the structure of intelligence. This has included a theory that there is a general (g) factor representing what all the tests have in common. Others have emphasized the specific group factors such as memory or verbal comprehension.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Intelligence and Intelligence Testing

  • Richard B Fletcher, John Hattie, Richard Fletcher(Authors)
  • 2011(Publication Date)
  • Routledge

    (Publisher)

  At the centre of the historical debate on IQ and its testing are arguments such as intelligence is immutable and intelligence testing is merely a sorting device for placing people into neat, identifiable categories from which they cannot escape. Antagonists such as Stephen Jay Gould (1981) have argued that the perniciousness of the IQ testing process means that a single score does not take into account the complexity of the individual and context in which they exist. Above all, Gould argued that IQ tests are simply a tool to maintain the social order by confirming a person’s societal position. After all, there are divisions in society and what better way to highlight these by using a seemingly ‘objective’ measure of ability. In other words, he argued IQ tests are blunt instruments with which to maintain societal order while at the same time allowing one to passively ignore the complex issues and problems that underlie differential levels of ability.

  If only the debate were so simple. Sure, it is easy to debunk the whole notion of intelligence and the associated methods of assessing it. It is easy to claim that ‘intelligence is what the intelligence test measures’ but that because measurement definitions are so narrow and culturally specific we should ban all intelligence testing. But such an argument is akin to throwing the baby out with the bath water. Most psychologists and psychometricians would agree that intelligence testing has its limitations, but most would also support the notion of intelligence as being a major dimension in an individual’s overall psychological make-up and, therefore, something that should not be ignored (as Wittgenstein commented, ‘The strength of the rope lies not in any one thread but in the overlapping of many fibres’ – intelligence is but one thread). Indeed, measures of intelligence can be important in many ways, as Chapter 4

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Reconstructing Educational Psychology

  • Bill Gillham(Author)
  • 2022(Publication Date)
  • Routledge

    (Publisher)

  6 THE FAILURE OF PSYCHOMETRICS DOI: 10.4324/9781003279853-6 Bill Gillham

  In the late seventies few psychologists give an intelligence test without a sense of unease, without a note of apology (or defensiveness) as if engaging in some shameful act. This is a most remarkable change and comparatively recent as a widespread phenomenon. Although partly a response to the Zeitgeist of liberal prejudice it is mainly due to an increasing awareness amongst psychologists of the conceptual and practical limitations of traditional tests.

  The arguments against the utility of intelligence tests have been around for a long time. Simon’s Intelligence Testing and the Comprehensive School was published in 1953 and, after twenty-five years, is still the best critique of psychometric concepts of intelligence. During the ensuing decade a number of important papers and books were published dismissing notions which had endured since the days of Lewis Terman — intelligence as ‘capacity’, intelligence as distinct from ‘attainments’, intelligence as a stable and relatively unteachable quality (Pidgeon and Yates, 1956; Vernon, 1958; Crane, 1959; Liverant, 1960; Hunt, 1961).

  Yet for a long time the practice of intelligence testing seemed little affected by these criticisms — one manifestation of the durability of powerful, simple and easily communicated ideas. As late as 1965 Holtzman, writing about the concept of intelligence, could say that ‘one of the most significant accomplishments of psychology has been the development of tests for measuring intelligence’ (Brim et al., 1966). Finishing my training as an educational psychologist in that year I am sure that I would have agreed with Holtzman. I can certainly remember being pleased to hear that a large-scale project was being set up at the University of Manchester to develop a British

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Educational Differences (RLE Edu L)

  • Arthur Jensen(Author)
  • 2012(Publication Date)
  • Taylor & Francis

    (Publisher)

  Life. I emphasize ‘present’, because this is my answer as of May, 1968 – not guaranteed to be perfectly correlated with my views on the subject six months or a year from now, although I would surely expect a substantial positive correlation, for we are not totally without bearings in this field. The current pace of relevant re-search, however, is such that anyone who hopes to view these issues constructively and creatively must assiduously eschew a doctrinaire stance.

  An old issue

  The issue of cultural bias or status bias in intelligence tests is as old as intelligence testing itself. Alfred Binet in 1905 made a clear distinction between the kinds of judgment, adaptability, and general problem-solving ability he called intelligence and attempted to measure by means of his mental age scales, on the one hand, and, on the other, the kinds of information acquired in schools or in a cultured home. Despite his efforts to come as close as possible to assessing the child’s innate endowment of general intelligence by means of his scales, he consistently found systematic differences between various social status groups. The first formal study of this social aspect of intellectual assessment was published by Binet just five years after the appearance of the first edition of his now famous intelligence test, which became the prototype of nearly all subsequent individual tests of intelligence (1916). Binet reported evidence from France and Belgium that children of professional workers did better on his new intelligence tests, on the average, than did children in working-class neighborhoods. Since then, the question of social-class bias in tests versus real social-class differences in intelligence has been an issue of dispute among psychologists, sociologists, and educators. Innumerable investigations have been made in the United States, in Europe, and in Asia, of the relationship of social status to performance on intelligence tests. These investigations have used a wide variety of intelligence tests and many different methods of measuring social status. Without a single exception, the studies show a positive correlation between intelligence test scores and social status; half of the studies yield correlations between 0·25 and 0·50, with a central tendency in the region of 0·35 to 0·40. When children selected from the total population are grouped into social status categories, the mean IQs of the groups differ by as much as one to two standard deviations (15 or 30 IQ points), depending on the method of status classification. The fact of social class differences in measured intelligence is thus about as solid a fact as any that we have in psychology, and apparently it has long since ceased being a point of dispute. Most of this evidence has been reviewed by Kenneth Eells et al.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  AP® Psychology All Access Book + Online + Mobile

  • Nancy Fenton, Jessica Flitter, Jessica Flitter(Authors)
  • 2015(Publication Date)
  • Research & Education Association

    (Publisher)

  Chapter 13 Testing and Individual Differences

  Our society believes that intelligence is very important; in fact, intelligence tests are used to decide what schools people may attend, what jobs they are qualified for, and whether or not they are capable or brilliant. Interestingly, psychologists generally cannot agree on what exactly intelligence is or an effective way that it can be measured. Experimental psychologist E.G. Boring once famously declared, “Intelligence is what intelligence tests measure.” If that were the only meaning of intelligence, no one but psychologists would care about it. Broadly defined, intelligence involves the capacity to acquire knowledge, reason effectively, and adapt to one’s surroundings by utilizing a combination of inherited abilities (nature) and learned experiences (nurture). Intelligence may be described as a collection of separate abilities or as one significant factor. The study of intelligence is closely related to the field of psychometrics , or the scientific study of using mathematical or numerical methods to measure psychological variables by creating reliable and valid tests. Psychometrics involves the use of a mathematical concept called factor analysis , which utilizes statistics to reduce the number of variables by placing them in clusters of related items. This technique can be used to determine groups of similar variables on a test to determine if an individual’s score on a measure of intelligence is related to one type of ability (factor) or many. In regard to intelligence, it is important that you are able to describe the main theories, the history of testing, and the elements of strong test construction.

  Intelligence Theories

  As psychologists have researched intelligence, a variety of different theories have been proposed to explain how intelligence can be defined and potentially measured. The theories differ in terms of how they explain the overall concept. Some theories describe intelligence as one type of ability; others describe intelligence as being comprised of several different abilities. The major theories of intelligence, both historical and contemporary, that are required for the AP Psychology exam are outlined in this chapter.

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Psychological Testing

  Principles, Applications, and Issues

  • Robert Kaplan, Dennis Saccuzzo(Authors)
  • 2017(Publication Date)
  • Cengage Learning EMEA

    (Publisher)

  In one survey, 1020 experts agreed that there were some sociocultural biases in the tests (Snyderman & Rothman, 1987). However, these experts also generally agreed that the tests were valid for Copyright 2018 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-300 CHAPTER 19 ● Test Bias 535 predictive purposes. Their main concerns involved the interpretation and application of test results by elementary and secondary schools. A more recent study surveyed opinions about psychological tests among experts from diverse communities, including different countries, regions, such as Finland, East Asia, sub-Saharan Africa, Southern Europe, the Arabian-Muslim world, Latin America, Israel, Jews in the West, Roma (gypsies), and Muslim immigrant. They found that education was the most commonly listed reason for international differences in cognitive ability. Genetics and cultural factors were identified as the second and third most commonly listed reasons, respectively. (Rindermann, Becker, & Coyle, 2016). In general, industrial and organizational psychologists tend to feel that ability testing does not discriminate by race. In one study of 703 members of the Society of Industrial and Organizational Psychology, there appeared to be consensus that cognitive ability tests are valid and fair. However, the I/O psychologists also felt that tests provide an incomplete picture of human abilities and that job selection should consider tests as only one component. Perhaps the most controversial defense of testing was presented in a 1994 book entitled The Bell Curve. This book is reviewed in Focused Example 19.4. Thinking Differently: Finding New Interpretations of Data Clearly, the observed differences between minority and nonminority groups on standardized tests pose a problem.

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Child Psychology

  Development in a Changing Society

  • Robin Harwood, Scott A. Miller, Ross Vasta(Authors)
  • 2012(Publication Date)
  • Wiley

    (Publisher)

  How does cultural bias in testing work to influence IQ scores? 4. What is stereotype threat and how might it influence IQ scores? Alternative Conceptions of Intelligence The traditional, psychometric definition of intelligence and its uses in predicting school performance has been challenged, leading to the development of alternative conceptual- izations of intelligence. Three influential alternatives are the evolutionary model, Stern- berg’s triarchic theory and dynamic testing, and Gardner’s theory of multiple intelligences. These alternatives do not focus primarily on abilities associated with success in school. School and the disciplines taught in school are recent developments in human history. Reading and writing trace back fewer than 10,000 years. To this day, in some societies the majority of citizens cannot read, and formal education is unknown. Consequently, many theorists have sought other ways to define intelligence. Evolutionary Approaches According to the evolutionary model, intelligence is the ability to adapt to the environment, and is a product of natural selection occurring over tens of thousands of years. Many of the TABLE 9.3 Risk Factors in a Study of Family Environment and IQ R ISK FACTOR D ESCRIPTION Occupation Head of household was unemployed or held unskilled occupation. Mother’s education Mother did not complete high school. Family size Family has four or more children. Father absence Father was not present in the home. Stressful life events Family experienced 20 or more stressful events during the child’s first 4 years. Parenting perspectives Parents held rigid or absolutist conceptions of children and childrearing. Maternal anxiety Mother was unusually high in anxiety. Maternal mental health Mother had relatively poor mental health. Mother-child interaction Mother showed little positive affect toward the child. SOURCE: Adapted from A.J. Sameroff, R. Seifer, A. Baldwin, and C.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Clinical Psychology

  The Study of Personality and Behavior

  • Max Gluckman, Sol L. Garfield(Authors)
  • 2017(Publication Date)
  • Routledge

    (Publisher)

  The problem of validity is much more difficult to answer satisfactorily. Again, it can be mentioned that this is not a criticism of the Stanford-Binet alone, but of all psychological tests of intelligence. Problems in defining intelligence and securing suitable criteria for test validation are some of the difficulties which face the psychologist in this area. Since a discussion of these complex problems would lead us too far astray from our main purpose, the reader is referred to other sources for information on these topics (Goodenough, 1946; Matarazzo, 1972; McNemar, 1964; Stoddard, 1943; Wechsler, 1958). It is important, however, to emphasize such matters to the beginning clinician. Too often the emphasis in clinical psychology has been on the giving and interpreting of psychological tests with little attention devoted to basic problems of standardization and validity. A competent clinical psychologist, however, must be familiar with all the important aspects of psychological testing, including some understanding of the variables to be appraised.

  The validity of the test items in the revised Stanford-Binet Scale was evaluated by noting the increase in the percentage of subjects passing each item at successive chronological and mental ages. In other words, since intelligence develops with age, one can assume, up to a certain point, that older children are more intelligent than younger ones. Valid test items, therefore, should be passed by a greater percentage of children at successive age levels. The authors of the scale readily state that this is not conclusive evidence of validity. But, in the absence of suitable criteria of intelligence, more direct evidence is difficult to secure. The careful and comparatively extensive work on the standardization of this scale, the clinical usefulness of the original Stanford-Binet, and the close agreement between the two scales contributed to the acceptance of the revised form.

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Personality and Individual Differences

  • Tomas Chamorro-Premuzic(Author)
  • 2015(Publication Date)
  • BPS Blackwell

    (Publisher)

  DISCUSSION Traditional notions of intelligence have always emphasized individual differences in the ability to adapt to the real world, though they rarely took into account how individuals may do this in unfamiliar environments. The concept of CQ presents an interesting avenue for research in this area. Source: Ang, S., Van Dyne, L., & Koh, C. (2005) Personality correlates of the four-factor model of cultural intelli- gence, Group Organization Management 31, 100–23. 5.3.6 An Overview of Intelligence Testing So far, we have investigated the history of intelligence testing and the primary intel- ligence tests which have emerged; but what characterizes a good intelligence test, and what are their limitations? Firstly, a good intelligence test will be standardized. That is, the conditions of administration will be controlled as much as possible so that, in theory, the only dif- ferential variable between tests is intelligence. The instructions, environment, and so on should be identical each time the test is administered. A good test should also be standardized in the sense that norms are identified across a population, in which comparisons can therefore be made. There are, however, a number of issues to consider with respect to intelligence testing. Although the internal reliability of intelligence tests is typically well estab- lished, the point may be moot since these tests are usually developed on the basis that different components should correlate with one another; if an item or facet does not correlate with the others, it may be dropped from the measure. More pressing is the issue of test-retest reliability. The same person taking the same test on two different days may obtain a different IQ score due to situational and state factors such as stress or distraction.

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Introduction to Clinical Psychology

  Bridging Science and Practice

  • Douglas A. Bernstein, Bethany A. Teachman, Bunmi O. Olatunji, Scott O. Lilienfeld(Authors)
  • 2020(Publication Date)
  • Cambridge University Press

    (Publisher)

  What additional evidence would help to evaluate the alternatives? If the problem of test bias really is a reflection of differences among various groups’ opportunities to develop their cognitive skills, it will be important to learn more about how to reduce those differences. Making cultures fairer by enhancing the skill development opportunities of traditionally disadvantaged groups should lead to smaller differences among groups on tests of cognitive ability (Martinez, 2000). At the same time, alternative tests of cogni- tive ability must also be explored, particularly those that include assessment of problem- solving skills and other abilities not measured by most intelligence tests (e.g., Shaunessy, Karnes, & Cobb, 2004; Sternberg & Kaufman, 1998). For example, there is growing evidence that standard IQ tests do an inadequate job of assessing the ability to think rationally, a cap- acity that is essential in many domains of everyday life (Stanovich, 2009). If new tests show smaller between-group differences than traditional tests, but have equal or better ability to predict people’s academic or occupational performance, many of the issues discussed in this section will have been resolved. So far, efforts in this direction have not been especially successful. What conclusions are most reasonable given the kind of evidence available? In short, at this stage, differences in the average intelligence test scores of various racial and ethnic groups appear more likely to be due to differences in cultural conditions than to unfair bias in the tests themselves, though both prob- ably contribute. The distinguished psychometri- cian Anne Anastasi summarized the situation this way: Tests are designed to show what an individual can do at a given point in time. They cannot tell us why… Tests cannot compensate for cultural deprivation by eliminating its effect from their scores.

  Sign up to read

  Learn more about book