Hans J. Eysenck`s Influence on Intelligence Research - CEON-a
Transcription
Hans J. Eysenck`s Influence on Intelligence Research - CEON-a
PSIHOLOGIJA, 1998, 3, 249-256 UDC 159.95.07 Hans J. Eysenck's Influence on Intelligence Research JOHN C. WICKETT Assessment Strategies, Ottawa1 Sir Francis Galton adopted a peculiar approach to measuring cognitive ability: he presumed intelligence to be observable in the physiological processing of the body. This Galtonian approach went unrewarded for many years, until Eysenck revived it by revisiting reaction time as a key variable. Out of this work grew Eysenck's belief that intelligence, and particularly g, is best described by a speed of information processing model. Recent study has focused on how this speed is manifest in the body. Nerve conduction velocity studies have shown that faster NCV predicts higher IQ, but does not correlate with RT, indicating a dissociation between mental and neural speed. Solving this paradox is just one area of future study. The biological approach, as resurrected and refined by Eysenck, is quickly supplanting (or perhaps explaining) cognitive models, as witnessed by the surge of research on brain volume, cerebral glucose metabolism, and quantitative trait loci. " . . . I can still feel the effects of great poetry on my autonomic nervous system." Hans J. Eysenck (1990), Rebel with a Cause The earliest days of intelligence research were dominated by a belief that cognitive ability was manifest in all aspects of the individual. Intelligence was a global trait that could be observed in the physical and the mental. Measurement of the ability to discriminate two weights was not seen as theoretically different from ———————— 1 50 Driveway, Ottawa, Ontario, K2P 1E2, Canada 249 J. C. Wickett measurement of the ability to discriminate two ideas. This was a view championed by Sir Francis Galton and James McKeen Cattell in the late 19th century. These early researchers made an incredible assumption: Galton and Cattell, among others, were positing a continuity and congruity between the physical and the mental that we still have a hard time reconciling. For these scientists, not only could the psychological be measured through the physical, but it actually was the physical. The turn of the century saw two relevant pieces of research get published. The first was the correlational analysis of Cattell's testing programme as conducted by Clark Wissler (1901). Wissler found no relation between Cattell's mental tests and the school standing of Columbia students. Furthermore, there was no evidence that the tests themselves intercorrelated. We know now that this analysis was fraught with reliability, restriction of range, and criterion problems, but at the time this was taken as strong evidence that an appeal to psychophysiological methods would not best serve the measurement of mental ability. This signaled the end of the Galtonian approach to intelligence measurement. Alfred Binet was soon to introduce his intelligence scale, and this forerunner to the modern IQ test quickly supplanted all others. For decades to come, the field's main advances were seen in the technological refinements of the scales, rather than in new understanding of the foundation of cognitive ability. The other relevant piece of research, and the primary exception to the preceding sentence, was the demonstration of general intelligence by Charles Spearman (1904). Taking the concept of g to its logical end suggests not only a communality among cognitive ability tests, but a communality between psychological performance and the physical architecture which provides that performance. Thus, in many ways, the existence of g is the fundamental proof that Galton was right. However, this was not to provide any real revival of Galtonian ability testing - this had run its course, and it was more important in these early days of intelligence research to provide a working intelligence test than one that was theoretically precise. Many years passed . . . Today, intelligence researchers are in agreement that there is a robust g factor, at the apex of a few group ability factors (Carroll, 1993; Jensen, 1998). It is now considered axiomatic that faster reaction time (RT) on elementary cognitive tasks is positively associated with greater IQ, and that the more variable the RTs the lower the IQ (Vernon, 1987). From these two now well-established findings has sprung a fount of research. The reaction time findings in particular lead to the conclusion that there are likely some basic, fundamental processes that underlie intelligence. It seems perfectly likely that this fundamental process would be observable in the body. That g shows itself so strongly suggests that these processes would be observed in many aspects of the body, and not just in one specific region. The field is now awash with studies aimed at determining what is this biological basis of intelligence. Researchers examine nerve conduction velocity, hormonal fluctuations, electroencephalographic waveforms, brain volume, and molecular genetics to determine what biological variables exert an influence on psychometric 250 Eysenck's Influence on Intelligence Research intelligence (see, for reviews, Deary & Caryl, 1997; Vernon, 1997; Vernon, Wickett, Stelmack, & Bazana, in press). We have come full circle. The methods are more advanced, but we are simply following theory that Galton conceptualized more than 100 years ago. Of course, this paper is not about Galton. Galton is recognized by many of those in the field as the father of intelligence testing (in the broadest sense). This is, at least in part, because his theoretical leanings and research protocols were in line with what we believe and practice today. The question has to be asked, though, whether we would see this influence to the extent that we do if it were not for the efforts of Hans J. Eysenck. It is contended here that it was Eysenck's recognition of a potentially fruitful line of study that has led in large part to the rebirth of biological approaches to understanding human cognitive diversity. It has been argued that the time was ripe for a change of this nature (Caryl, 1997), and that it may well have happened without Eysenck's particular influence. There is probably a good deal of truth to this view - nonbiologically-based approaches to understanding intelligence were not yielding much in the way of discovery, and so change was inevitable. Perhaps the most accurate statement that can be made about Eysenck's influence is that he provided an atmosphere that was strongly supportive of a new biological approach, and gave courage to those who may have been daunted by the prospect of tackling controversial and unpopular problems. Intelligence was not the primary focus of Eysenck's energies, but it was the first - at least in terms of publication. His first journal article (Eysenck, 1939) was a factor analysis (using Cyril Burt's method) of Thurstone's Primary Mental Abilities, where he found strong evidence for g, beyond the set of primary factors. In his autobiography, Eysenck (1990) mentions this as one of his first experiences with odd behaviour exhibited by Burt. Eysenck performed the calculations, and Burt wrote the text (and did so such that highly favourable light was cast upon himself), but only Eysenck's name appeared as author. Further, the text that Burt provided was different from that which he had originally shown to Eysenck. Disregarding the oddities surrounding this publication, it has two main distinctions which make it relevant. First, it demonstrated for Eysenck that g was important. This was to later play a major role in his theorizing. Second, although both this earliest research and his Ph.D. advisor were firmly within the intelligence area, he was to not place particular focus on intelligence for almost 30 years, favouring his work on personality instead. The exact cause of the decision can only be speculated upon, but Jensen (1997) suggests that Burt's overpowering role in the intelligence field may have led Eysenck to choosing another, less-traveled path. Also, the behaviour of Burt may have left something of a bad taste, making the personality path more favourable by comparison. Eysenck fully entered the intelligence arena with the publication of his call for a unification of theory and empiricism in the field in 1967 (although this was presaged in some ways in Eysenck, 1953). Since this time, his industry and influence have been monumental, as evidenced in the tribute to Eysenck edited by Helmuth Nyborg (1997). By way of disclaimer, what is presented here makes no claims at 251 J. C. Wickett comprehensiveness, but is rather a small smattering of what has been given much fuller treatment elsewhere. Even a cursory examination of Eysenck's publications quickly informs the reader that one of his main objectives was the revival of a Galtonian approach to understanding intelligence. Numerous theoretical articles, chapters, and books were framed as doing exactly this (Eysenck, 1967, 1979, 1988, 1995). He dedicated his 1979 book, The Structure and Measurement of Intelligence, to the legacy of Galton. Most important, of course, his stress on the importance of using elementary indices of intelligence, and of intelligence ultimately being based in biology, is entirely consistent with Galton's work. Although tracing back lines of influence from Eysenck to Galton requires little effort, this is not overly compelling as explanatory in how Eysenck came to adopt his views. What appears more likely is that Eysenck simply recognized this approach as one that could have merit (subject, of course, to empirical support). That he decided to give it more consideration when he did can be traced to his reading of a German article by Roth (1964) which addressed the relation between RT and IQ. This study provided empirical support for Eysenck's views, and demonstrated a clear direction for future research. The years that followed saw much work directed at understanding the ways in which reaction time related to intelligence, with much of this work being conducted by Arthur Jensen upon Eysenck's prompting. As it became clear that faster and more consistent performance on elementary cognitive tasks was reliably associated with greater intelligence, a theory began to take hold in the field. Eysenck's (1967) speed of information processing model suggested, quite simply, that fast and error-free processing was what allowed for greater intellectual feats. An individual who can process information more quickly can process more information because less will drop out of limited-capacity working memory. The intervening years have not seen great elaboration to this model, but it cannot be said that it has not served the field well. The logical question becomes one of what is the biological manifestation of 'mental speed'. This question took Eysenck down two main paths. First, he considered the relation between various EEG variables and IQ. Eysenck placed particular focus on the Hendricksons' string measure paradigm, finding little support for the early highly-positive results (although several theory-consistent results were obtained with other EEG variables; Eysenck, 1979, 1983; Barrett & Eysenck, 1992; Bates & Eysenck, 1993). Undaunted, Eysenck examined an even purer measure of neural functioning - nerve conduction velocity along the median nerve of the arm. Barrett, Daum, and Eysenck (1990) and Barrett and Eysenck (1993) report two studies of nerve conduction velocity. In neither study do the authors find much support for a relation with IQ, but others have. Vernon et al. (in press) review this literature, demonstrating that faster nerve conduction velocity is associated with greater IQ. As with many areas of electrophysiological research, all studies do not tend to converge on a definitive estimate of the magnitude (or even existence) of 252 Eysenck's Influence on Intelligence Research this effect. Vernon and Mori (1992) report correlation coefficients above .40; Wickett and Vernon (1994) found essentially no effect; and Tan (1996) has even found a -.61 correlation in females. Other studies have tended to come down on the side of either a modest positive correlation, or no effect at all. As Vernon et al. (in press) report, the mean effect is for about a .30 correlation in males, but no effect in females. The most perplexing aspect of this line of research (aside from the fact that neural conduction speed along the arm is correlated with scores on an IQ test) is that no correlation between velocity and reaction time exists. So, mental speed and neural speed are not the same thing, and in fact the one has nothing to do with the other. The biological basis of the RT-IQ relation thus remains elusive. In addition to nerve conduction velocity and average evoked potentials, several other biological variables are the focus of study in current-day research. Vernon (1997) and Vernon et al. (in press) review this literature, in particular examining the developments in relation to brain volume, cerebral glucose metabolism, and quantitative trait loci. Findings in each of these areas have indicated strong biological underpinnings to intelligence. By all appearances, this Galtonian/Eysenckian approach has taken firm hold of the field, and will shape it for many years to come. The phenomenal rate of technological advance in the areas of imaging and molecular genetics means that new biological variables are continuously presenting themselves as potential targets of study. The next few years promise monumental discoveries. References Barrett, P. T., Daum, I., & Eysenck, H. J. (1990) Sensory nerve conduction and intelligence: A methodological study. Journal of Psychophysiology, 4, 1-13. Barrett, P. T., & Eysenck, H. J. (1992) Brain evoked potentials and intelligence: The Hendrickson paradigm. Intelligence, 16, 361-381. Barrett, P. T., & Eysenck, H. J. (1993) Sensory nerve conduction and intelligence: A replication. Personality and Individual Differences, 15, 249-260. Bates, T. C., & Eysenck, H. J. (1993) String length, attention and intelligence: Focused attention reverses the string length-IQ relationship. Personality and Individual Differences, 15, 363-371. Carroll, J. B. (1993) Human Cognitive Abilities. Cambridge: Cambridge University Press. Caryl, I. J. (1997) Intelligence and information processing. In H. Nyborg (Ed.), The Scientific Study of Human Nature: Tribute to Hans J. Eysenck at Eighty. Oxford: Elsevier Science Ltd. 253 J. C. Wickett Deary, I. J., & Caryl, P. G. (1997) Neuroscience and human intelligence differences. Trends in Neurosciences, 20, 365-371. Eysenck, H. J. (1939) Primary mental abilities. British Journal of Educational Psychology, 9, 270275. Eysenck, H. J. (1953) Uses and Abuses of Psychology. London: Pelican. Eysenck, H. J. (1967). Intelligence assessment: A theoretical and experimental approach. British Journal of Educational Psychology, 37, 81-98. Eysenck, H. J. (1979) The Structure and Measurement of Intelligence. Berlin: Springer-Verlag. Eysenck, H. J. (1983) Psychophysiology and intelligence: New methods of measuring the I. Q. Indian Journal of Psychophysiology, 1, 33-39. Eysenck, H. J. (1988) The biological basis of intelligence. In S. H. Irvine & J. W. Berry (Eds.), Human Abilities in Cultural Context. New York: Cambridge University Press. Eysenck, H. J. (1990) Rebel with a Cause. London: W. H. Allen. Eysenck, H. J. (1995) Can we study intelligence using the experimental method? Intelligence, 20, 217-228. Jensen, A. R. (1997) Introduction: Hans Eysenck and the study of intelligence. In H. Nyborg (Ed.), The Scientific Study of Human Nature: Tribute to Hans J. Eysenck at Eighty. Oxford: Elsevier Science Ltd. Jensen, A. R. (1998) The g Factor. Westport, CT: Praeger. Nyborg, H. (Ed.). (1997) The Scientific Study of Human Nature: Tribute to Hans J. Eysenck at Eighty. Oxford: Elsevier Science Ltd. Roth, E. (1964) Die Geschwindikeit der Verarbeitung von Information und ihr Zusammenhang mit Intelligenz. Zeitschrift fur Experimentelle und Angewandte Psychologie, 11, 616-622. Spearman, C. (1904) "General Intelligence" objectively determined and measured. American Journal of Psychology, 15, 201-293. Tan, U. (1996) Correlations between nonverbal intelligence and peripheral nerve conduction velocity in right-handed subjects: Sex-related differences. International Journal of Psychophysiology, 22, 123-128. Vernon, P. A. (1987) Speed of Information Processing and Intelligence. Norwood, NJ: Ablex. Vernon, P. A. (1997) Behavioral genetic and biological approaches to intelligence. In H. Nyborg (Ed.), The Scientific Study of Human Nature: Tribute to Hans J. Eysenck at eighty. Oxford: Elsevier Science Ltd. Vernon, P. A., & Mori, M. (1992) Intelligence, reaction times, and peripheral nerve conduction velocity. Intelligence, 16, 273-288. 254 Eysenck's Influence on Intelligence Research Vernon, P. A., Wickett, J. C., Stelmack, R. M., & Bazana, P. G. (in press) The neuropsychology and psychophysiology of human intelligence. In R. J. Sternberg (Ed.), Handbook of Human Intelligence (2nd ed.). Cambridge: Cambridge University Press. Wickett, J. C., & Vernon, P. A. (1994) Peripheral nerve conduction velocity, reaction time, and intelligence: An attempt to replicate Vernon and Mori (1992) Intelligence, 18, 127-131. Wissler, C. (1901) The correlates of mental and physical tests. Psychological Review, Monograph No. 3. Uticaj Hansa Ajzenka na istraživanje inteligencije DŽON C. VIKET Ser Frensis Golton usvojio je osoben pristup merenju kognitivne sposobnosti: pretpostavio je da se inteligencija može posmatrati preko fizioloških procesa u telu. Goltonovski pristup ostao je nepriznat tokom mnogih godina, sve dok ga Ajzenk nije oživeo uzimajući vreme reakcije kao ključnu varijablu. Iz tog rada proisteklo je Ajzenkovo uverenje da se inteligencija, a naročito g, može najbolje opisati modelom brzine obrade informacija. Novije istraživanje usmereno je na to kako se ova brzina manifestuje u telu. Istraživanja brzine nervnog provođenja (NCV) pokazala su da brže NCV prognozira viši IQ, ali da ne korelira sa vremenom reakcije, što ukazuje na razdvajanje mentalne i neuralne brzine. Rešavanje ovog paradoksa je samo jedna oblast za buduća istraživanja. Biološki pristup koji je Ajzenk ponovo uveo i prečistio, brzo potiskuje (ili možda objašnjava) kognitivne modele, o čemu svedoči veliki talas istraživanja o volumenu mozga, metabolizmu cerebralne glukoze i lokusu kvatitativne crte. Vli®nie Hansa AŸzenka na issledovanie intellekta D@ON C. VIKET Sƒr Fransis Gol√ton razrabotal svoeobrazn∫Ÿ metod dl® izmereni® kognitivnoŸ sposobnosti: on predpolagal, ~to intellekt mo`no nablÓdat√ ~erez fiziologi~eskie process∫ v tele. Ego podhod ostavals® nepriznann∫m na prot®`enii mnogih let, poka AŸzenk ne o`il ego, opredeliv vrem® reakcii kak klÓ~evuÓ peremennuÓ. Na osnovanii ƒtoŸ idei AŸzenk pri{el k v∫vodu, ~to intellekt i, v ~astnosti g mo`no obÍ®snit√ lu~{im sposobom primenneniem modeli skorosti obrabotki informaciŸ. Dal√neŸ{ie 255 J. C. Wickett issledovani® napravlen∫ na opredelenie manifestacii vremennogo uslovi® v tele. Issledovani® skorosti provedeni® nervnogo vozbu`deni® (NCV), pokazali, ~to bolee b∫str∫Ÿ NCV prognoziruets® bolee v∫sokim IQ, no ne so~etaets® s vremenem reakcii. Ïto obsto®tel√stvo ukaz∫vaet na razdvoennost√ umstvennoŸ i neŸral√noŸ skorosteŸ. Razre{enie ƒtogo paradoksa ostaets® odtel√noŸ oblast√Ó dl® buduçih issledovaniŸ. Biologi~eskiŸ podhod, zanovo vvedenn∫Ÿ AŸzenkom, b∫stro poddavl®et (ili, vozmo`no, obÍ®sn®et) kognitivn∫e modeli, o ~em svidetel√stvuet volna issledovaniŸ obÍema golovnogo mozga, metabolizma cerebral√noŸ glÓkoz∫ i lokusa koli~estvennoŸ linii. 256