grady@Apple.COM (Grady Ward) (03/21/88)
From "The Structure and Measurement of Intelligence" by Hans J. Eysenck, Ph.D., D. Sc. and David W. Fulker, Ph.D. (University of London, Institute of Psychiatry, De Crespigny Park, Denmark Hill, London SE5 8AF) Springer-Verlag 1979, pp. 227-228 (Epilogue): It is possible, and may be useful, to pull together the major conclusions to which the research summarized and surveyed in this book may lead us. These conclusions are of course not final verities, never to be challenged; they are simply points of view, concepts and generalizations which are supported by the existing evidence, but which might be disproved by experiments in the future. Taking into account this uncertain status (which of course is shared to a greater or lesser degree by all scientific findings), we may list our major conclusions as follows: 1. Cognitive, intellectual performance can be described objectively in terms of concepts like abilities, intelligence, etc. 2. It is possible to measure these concepts empirically, through the use of tests, problems, and questions, and indirectly through psychophysiological measures. 3. The intercorrelations between special types of tests define certain primary mental factors, usually referred to as primary or group factors. 4. Primary abilities correlate together to form higher order concepts, such as fluid and crystallized intelligence; these concepts are defined in terms of observed correlations. 5. Underlying such higher order concepts as "intelligence" there are physiological structures, such as those giving rise to individual differences in evoked potentials. 6. Among major primary factors or abilities thus far demonstrated by research are verbal ability, numerical ability, memory, perceptual ability, divergent ability, reasoning, visio-spatial ability, and several others. 7. The IQ, as measured by standard tests, can be shown to be a product of the three major independent factors; mental speed, persistence, and error checking mechanisms. This "splitting" of the IQ suggests more analytical research designs than have been customary hitherto. 8. Intelligence as measured by IQ tests has a strong genetic basis; genetic factors account for an estimated 80% of the total variance, although this estimate has a standard error of some 5% to 10% attached to it. 9. Intelligence as measured by IQ tests is markedly affected by environmental factors; such factors account for an estimated 20% of the total variance, but this estimate too is of course subject to a sizeable standard error. 10. Genetic factors in intelligence are largely additive, but with a demonstrable contribution from assortative mating and dominance, This accounts for the fact that the broad heritability is some 10 points higher than the narrow heritability. 11. Primary abilities, when the influence of general intelligence is removed, also show evidence of genetic determination, sometimes with sex linkage (visuo-spatial ability) suspected. 12. Intelligence, as measured by IQ tests, is related to social behaviours, especially as evidence in educational achievement, determination of social class, and income (earnings). These in turn show evidence of genetic determination, mediated in part through differences in intelligence. 13. The facts summarized above lead to a definition of general intelligence as general, inherited mental ability; IQ tests measure this ability only with a certain degree of inaccuracy, being influenced to a variable extent by environmental factors. 14. The paradigm here outlined is entirely quantitative and scientific; there has been no attempt to deal with the so-called humanistic and other idiographic (subjective) ways of attacking the problem. Approaches such as Piaget can easily be accommodated within the present scheme. 15. The concepts used in this book are human inventions, like all scientific concepts; they are abstractions which cannot be said to exist or not to exist. Concepts are useful or useless; they cannot be true or false. It is only by this criterion that the theories here discussed should be judged. (This work is subject to copyright. "Fair use," as defined by the 1976 Copyright Act, is claimed.) --------------------------------------------------------------------------------- If anyone has facts to dispute the above claims, please cite them. Thank you. --Grady Ward
hollombe@ttidca.TTI.COM (The Polymath) (03/24/88)
In article <7742@apple.Apple.Com> grady@apple.UUCP (Grady Ward) writes: >From "The Structure and Measurement of Intelligence" by Hans J. >Eysenck, Ph.D., D. Sc. and David W. Fulker, Ph.D. (University of >London, Institute of Psychiatry, De Crespigny Park, Denmark Hill, London >SE5 8AF) Springer-Verlag 1979, pp. 227-228 (Epilogue): > >It is possible, and may be useful, to pull together the major conclusions to >which the research summarized and surveyed in this book may lead us. >These conclusions are of course not final verities, never to be challenged; >they are simply points of view, concepts and generalizations which are >supported by the existing evidence, but which might be disproved by >experiments in the future. Taking into account this uncertain status (which >of course is shared to a greater or lesser degree by all scientific findings), >we may list our major conclusions as follows: [etc., see root article] >If anyone has facts to dispute the above claims, please cite them. Thank you. > >--Grady Ward Eysenck was, at one time, famous for presenting what became known as "Eysenck's Challenge" to the psychological community. In it he purported to show, with statistical evidence, that psychotherapy with neurotics was useless and no more curative than no therapy at all. Eysenck's challenge was taken up and soundly refuted. Much of the refutation rested on pointing out that Eysenck had fudged his statistics. He had arranged the sampling and categories of his data to make the results fit his expectations. A more even handed evaluation of the same data produced quite different results. (Source: Korchin, S. J., _Modern Clinical Psychology_, New York: Basic Books, Inc., 1976, pages 428 - 432). Given the above, any statistical claims by Eysenck must be considered suspect without independent expert examination of the methodology used to obtain and verify them. After I've read his book, if I ever get around to it, I'll let you know what I think of his conclusions. "You can lie with statistics, but not to a statistician." -- B. Tabachnik, Ph.D. -- The Polymath (aka: Jerry Hollombe, hollombe@TTI.COM) Illegitimati Nil Citicorp(+)TTI Carborundum 3100 Ocean Park Blvd. (213) 452-9191, x2483 Santa Monica, CA 90405 {csun|philabs|psivax|trwrb}!ttidca!hollombe
rmpinchback@dahlia.waterloo.edu (Reid M. Pinchback) (04/09/88)
In article <7742@apple.Apple.Com> grady@apple.UUCP (Grady Ward) writes: >From "The Structure and Measurement of Intelligence" by Hans J. >Eysenck, Ph.D., D. Sc. and David W. Fulker, Ph.D. (University of >London, Institute of Psychiatry, De Crespigny Park, Denmark Hill, London >SE5 8AF) Springer-Verlag 1979, pp. 227-228 (Epilogue): > ... stuff >3. The intercorrelations between special types of tests define certain >primary mental factors, usually referred to as primary or group factors. > ... more stuff >8. Intelligence as measured by IQ tests has a strong genetic basis; genetic >factors account for an estimated 80% of the total variance, although this >estimate has a standard error of some 5% to 10% attached to it. > >9. Intelligence as measured by IQ tests is markedly affected by >environmental factors; such factors account for an estimated 20% of the total >variance, but this estimate too is of course subject to a sizeable standard >error. > >10. Genetic factors in intelligence are largely additive, but with a >demonstrable contribution from assortative mating and dominance, This >accounts for the fact that the broad heritability is some 10 points higher than >the narrow heritability. > >11. Primary abilities, when the influence of general intelligence is removed, >also show evidence of genetic determination, sometimes with sex linkage >(visuo-spatial ability) suspected. > >12. Intelligence, as measured by IQ tests, is related to social behaviours, >especially as evidence in educational achievement, determination of social >class, and income (earnings). These in turn show evidence of genetic >determination, mediated in part through differences in intelligence. > .... more stuff >If anyone has facts to dispute the above claims, please cite them. Thank you. > >--Grady Ward Most of it is in the form of neutral commentary, and critical analysis of an informal literature review 9 years old isn't too meaningful. To dispute empirical evidence in psychology (as in most sciences) generally requires examination of the original work (including nasty unpublished details that researchers tend to leave out). Having said that, the figures quoted (80/20 split on nature/nurture), are, I believe, the WITHIN groups differences. They aren't the BETWEEN groups difference figures. Comparing groups that differ in culture, sex, race, geographic location, etc., is a process that has come under attack, unless the relevant distinction is made. Thus a variable like inheritance is better at explaining IQ differences between your brother/sister and your cousins, than it is for explaining black/white IQ differences. The role of inheritance in explaining between-groups differences appears to be a much weaker one than it is for within groups. So, in particular, comment (12) is somewhat questionable. Literature references are Layzer[72], Lewontin[76], Loehlin, Lindzey & Spuhler[75], Scarr & Carter-Saltzman[82]. At this point in time, research results seem to be equivocal. Many confounds have been proposed. Perhaps a comprehensive review has been done more recently to include the results of recent research, I don't know. Maybe somebody out there in netland does. Reid M. Pinchback -----------------