Flynn Blair

of 68

  • date post

    30-May-2018
  • Category

    Documents

  • view

    217
  • download

    0

Embed Size (px)

Transcript of Flynn Blair

  • 8/14/2019 Flynn Blair

    1/68

    The history of intelligence

    New spectacles for developmental psychology

    James R. Flynn

    Clancy Blair

    The best-known developmentally based theory of intelligence is that of the Swiss

    psychologist Jean Piaget. He had little interest in individual differences. He illuminated

    how intelligence developed, in all children, from initial cognitive structures to later ones.

    These structures assimilated information and evolved to accommodate new information.

    Piagets tests were designed to record the development of intelligence. For example, to

    assess whether a child understood that the quantity of something was not altered by

    change of shape, water was poured from a shallow beaker to a tall one and the child

    asked whether there was more water than before.

    The psychometric approach to the study of intelligence did not arise out of

    developmental concerns but rather, focused on measuring intelligence differences

    between individuals. Charles Spearman was the first to give it a theoretical foundation.

    He identified the general factor of intelligence org. Its justification was empirical.

    Individual performance across a wide range of cognitive tasks tends to be positively

    correlated. Someone superior to the average person in classifying objects will tend to be

    superior in identifying the missing piece of a matrix and so forth.

    1

  • 8/14/2019 Flynn Blair

    2/68

    However, right from the start, the psychometric tradition recognized that the

    development of intelligence had a sociological dimension. Every individual is a member

    of a group. For example, even within one nation, the social context in which a black

    child develops may differ profoundly from that of a white child. It was inevitable that

    scholars like Arthur Jensen would study group differences. The group differences of

    greatest interest were the general skills to which the label intelligence is usually applied

    and the two most important academic skills, namely, reading and mathematics.

    We will try to show that writing cognitive history integrates the developmental

    and psychometric approaches and therefore, sheds new light on cognitive abilities and

    their development. To make our case, we will do seven things: discuss the psychometric

    and developmental traditions; summarize the cognitive history of the 20th century in

    America; distinguish cognitive history from the science of measuring intelligence;

    illustrate how cognitive history can correct mistakes in applied psychology that are

    difficult to self-correct; use it to clarify the roles of genes and environment; use it to show

    that enhancing intelligence is a life time quest; show how it revolutionizes the task of

    accounting for group differences.

    The psychometric and developmental traditions

    In combination, the psychometric (or differential) and developmental traditions cross-

    fertilize and provide valuable information about intelligence. But they begin from very

    different starting points and with different sets of assumptions. This is seen most clearly

    in the idea that intelligence within the psychometric tradition does not really develop per

    se. The emphasis is on the measurement of an underlying individual difference

    2

  • 8/14/2019 Flynn Blair

    3/68

    characteristic, one that is manifest in a variety of ways at different ages but is for the most

    part stable from very early until very late in the lifespan (Deary, 2000).

    The psychometric tradition recognizes that performance on the indicators used to

    assess intelligence improves from childhood to adulthood, for example, the size and

    extent of vocabulary knowledge, an indicator of what is referred to as crystallized

    intelligence, and the ability to complete increasingly complex puzzles or patterns, an

    indicator of what is referred to as fluid intelligence. But it does not treat this

    development as equivalent to enhanced intelligence. Rather it posits a stable underlying

    individual difference characteristic (IQ) and therefore, ranks the individuals scores on

    tests like vocabulary and analyzing patterns by comparing them to the scores of the same

    age individuals in the population from which that individual was drawn. These rankings

    (deviation IQ scores) tend to have excellent psychometric properties. They are internally

    very coherent, are remarkably similar from one time point to the next, and do about as

    good a job as any other available measure in providing information about an individual's

    level of competence relative to others, making them highly useful although rarified

    indicators of psychological functioning.

    This emphasis on stability over time and between-person (inter-individual)

    variability in the measurement of intelligence in the psychometric tradition stands in

    sharp contrast to an emphasis on change and within-person (intra-individual) variability

    in the developmental tradition. In the developmental tradition, the emphasis is on the rate

    and extent of change in the various indicators of intelligence, such as vocabulary or

    pattern completion, and most importantly on the determinants of change (Ferrer &

    3

  • 8/14/2019 Flynn Blair

    4/68

    McArdle, 2004). Rates of change relative to group averages and inter-individual variation

    are, of course, considered, but the primary interest is in the process of development, in

    contexts and experiences and aspects of the individual that affect intelligence.

    As a consequence, information about change and determinants of change in

    mental abilities in the developmental tradition leads to questions about intelligence that

    are essentially distinct from those of the psychometric tradition, which views intelligence

    as a fixed entity attributable to a specific cause or causes, such as the ability to process

    information quickly (Jensen, 1998) or the size and density of synaptic connections in the

    brain (Garlick, 2002). In contrast, the developmental tradition views intelligence as a

    malleable entity that is shaped during the course of development. Therefore, findings

    from the two approaches to the study of intelligence sometimes seem at odds if not

    antagonistic.

    New habits of mind

    In distinguishing the psychometric and developmental traditions, nothing brings them

    into relief quite as clearly as the phenomenon of rising mean IQ. The identification of

    population gains across successive cohorts of test takers on measures of fluid

    intelligence, with only modest change in crystallized intelligence (at least over the last 60

    years) amounts to one of the most surprising and challenging empirical findings of

    psychological research in the twentieth century (Flynn, 1987). Until recently, the notion

    of a unitary intelligence subject to the glacial pace of brain evolution held sway, with the

    corollary that dysgenic reproduction was slowly eroding cognitive potential. Massive IQ

    gains over time called this into question. Since the US data is most complete, we will

    4

  • 8/14/2019 Flynn Blair

    5/68

    make frequent use of it, but most of what we say applies to the West in general

    First, the gains have been huge. Since 1947, American children have gained 24

    points (SD = 15) on the Similarities subtest of the WISC (Wechsler Intelligence Test for

    Children). Estimated gains on Raven's Progressive Matrices, the standard measure of

    fluid intelligence, are even greater. Legitimate projections back to 1900, based on birth

    dates, put 20th century gains on these two tests at 50 points (Flynn, 2009, p. 23). The

    century saw Full Scale IQ gains of at least 30 points. On the other hand, gains were

    differential. Excellent data since 1947 show that gains on WISC subtests like

    Information, Arithmetic, and Vocabulary have been slight. We now know that society can

    make differential progress without facing some kind of psychometric veto from theg

    (general intelligence) factor.

    -------------------------------

    Insert Figure 1 about here

    -------------------------------

    The following traits allow us to solve problems with cognitive content: mental

    acuity, habits of mind, attitudes, knowledge and information, efficiency of information

    processing, and working memory. "Habits of mind" may need some explanation. Ten

    years ago, one of us began to do crossword puzzles. He now does them much better, but

    not because of increased mental acuity or even larger vocabulary or store of information.

    The usual proclivity with words is to use them to say something as simply and directly as

    possible. Crosswords require that this habit be modified to imagine secondary meanings,

    5

  • 8/14/2019 Flynn Blair

    6/68

    less literal meanings, reflect whether the clue word was being "used" as a noun or a verb,

    and so forth. At first, it was an effort to do all of this. Now it has become automatic -- a

    new habit of mind has replaced the old.

    The inculcation of habits of mind, of making previously effortful thinking

    automatic, makes reference to what are known as dual processing theories in cognitive

    psychology. Dual-processing theories define the general hypothesis that there are (at

    least) two primary ways in which the mind processes information. The first is relatively

    fast, automatic, and effortless and may involve non-conscious information processing to

    direct atte