National differences in intelligence and educational attainment

Intelligence 35 (2007) 115 – 121 National differences in intelligence and educational attainment Richard Lynn a,⁎, Jaan Mikk b a University of Ulster, Coleraine, Ireland b Šiauliai University, Lithuania Received 9 March 2006; received in revised form 7 June 2006; accepted 7 June 2006 Available online 7 July 2006 Abstract We examine the correlations between the national IQs of Lynn and Vanhanen (Lynn, R. and Vanhanen, T. (2002). IQ and the wealth of nations. Westport, CT: Praeger. Westport, CT: Praeger, Lynn, R. and Vanhanen, T. (2006). IQ and global inequality. Athens, GA: Washington Summit Books.) and educational attainment scores in math and science for 10- and 14-year olds in 25 countries and 46 countries (respectively) given in the TIMSS 2003 reports. It was found that national IQs had (attenuation corrected) correlations of between 0.92 and 1.00 with scores in math and science. The results are interpreted as a validation of the national IQs. They suggest that national differences in educational attainment may be attributable to differences in IQ, or alternatively that national IQs and in educational attainment are both indicators of the mental ability of national populations. It is also shown that national IQs are positively associated with national per capita income (r = .61). It is proposed that these have a reciprocal positive feedback relationship such that each augments the other. © 2006 Elsevier Inc. All rights reserved. Keywords: National IQs; Educational attainment; TIMSS 1. Introduction This paper reports high across country correlations between the national IQs calculated by Lynn and Vanhanen (2002, 2006) and national differences in educational attainment in mathematics and science. The paper has three objectives. First, to show that these correlations provide some validation of the national IQs. Second, to discuss the mechanisms underlying these correlations. And third, to examine the relationships between national differences in IQ and educational attainment and a number of educational, economic and demographic variables. ⁎ Corresponding author. University of Ulster, Coleraine, Ireland. E-mail address: [email protected] (R. Lynn). 0160-2896/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.intell.2006.06.001 The first objective is addressed because of the controversy concerning the validity of the national IQs. Among those who have accepted the national IQs as valid, it has been argued that they are a significant correlate of economic growth (Weede & Kämpf, 2002; McDaniel & Whetzel, 2004; Dickerson, 2006; Jones & Schneider, 2006; Whetzel & McDaniel, in press), measures of education (Barber, 2005), rates of suicide (Voracek, 2004), and skin color and winter temperature (Templer & Arikawa, 2006). Others, however, have contended that the national IQs presented by Lynn and Vanhanen (2002) have no validity. Among these, Ervik (2003, p.406) has written: “the authors fail to establish the reliability …of intelligence (IQ) test scores” (a measure that has no reliability must necessarily have no validity). Volken 116 R. Lynn, J. Mikk / Intelligence 35 (2007) 115–121 (2003, p.411) criticises the national IQs on the grounds of the “highly deficient data…in the type of IQ test used”. Barnett and Williams (2004, p.392) criticise the IQs on the grounds that the samples “are, in many cases, not representative of the countries from which they are derived” and assert that the cross-country “comparisons are virtually meaningless”. Hunt and Sternberg (2006, pp. 133,136) also assert that “the estimates of national IQ are technically inadequate” and “the concept of national IQ is meaningless”. The classical method of establishing the validity of intelligence tests is to show that they are correlated with educational attainment: “thousands of studies have been published, in numerous languages throughout the world, attempting to demonstrate the validity of intelligence tests against academic performance in school” (Matarazzo, 1972, p.281). Here we apply the same method to establish the validity of national IQs by showing that they are highly correlated with scores on tests of mathematics and science. The second objective of the present paper is to discuss the mechanisms underlying the correlations between national IQs and educational attainment in mathematics and science. National differences in educational attainment in mathematics and science have been reported since the 1960s in a series of reports of the International Mathematics and Science Studies (Husen, 1967; Comber & Keeves, 1973; Baker & Jones, 1993; Beaton et al., 1996a,b; Beaton et al., 1996a,b). The national differences in average scores in these studies have been quite consistent. In general, the Chinese, Japanese and Koreans have achieved the highest scores, followed by the Europeans, while a number of developing countries have had the lowest scores. It has not proved possible to find a full explanation for these differences. Several hypotheses have been examined, including those that the differences may be partly determined by expenditure on schools, class size, and the qualifications of teachers, but none of these have produced substantial associations with educational attainment. Although it is forty years since the publication by Husen (1967) of the first study of national differences in educational attainment in mathematics, the possibility that these might be associated with differences in intelligence, or even partly caused by differences in intelligence, has never been considered. The reason for this is that there have been no data on national differences in intelligence that could be used to examine these possibilities. This situation has changed with the publication by Lynn and Vanhanen (2002, 2006) of IQs for all nations in the world. This makes it possible to test the hypothesis, and this is the second objective of the present paper. The hypothesis that the national differences in educational attainment are associated with differences in intelligence is derived from the numerous studies in many countries showing that intelligence is associated with educational attainment at a magnitude of a correlation of around .5 to .7. The results of nine studies showing this are summarized in Table 1. This is not a complete list of such reports, but we believe it is typical. The first column gives the country, the second column gives the number in the sample, the third gives the age at which intelligence was measured, the fourth gives the age at which educational attainment was measured, and the fifth gives the educational subjects assessed by tests. The correlations range between .45 and .74 with a Table 1 Correlations between Intelligence and Educational Attainment Country Number Age IQ Age ed. att. Subject r Reference 1 2 Canada England 208 85 13 5 13 16 Several English .55 .61 3 England 85 5 16 Math .72 4 5 Britain N. Ireland 20,000 701 11 16 16 16 GCSE GCSE .74 .65 6 USA – – – Several .71 7 USA 455 13 13 Reading .68 8 USA – 18 18 Math .66 9 Switzerland 82 11 11 Math .45 Gagne and St. Pere, 2002 Yule et al., 1982 Yule et al., 1982 Deary, 2004 Lynn et al., 1984 Walberg, 1984 Lloyd and Barenblatt, 1984 Lubinski and Humphreys, 1996 Tewes, 2003 117 R. Lynn, J. Mikk / Intelligence 35 (2007) 115–121 median of .66. It makes little difference to the size of the correlations whether intelligence is measured early in childhood or in adolescence. One of the highest correlations (.72) is between intelligence measured at the age of 5 years and educational attainment in mathematics at the age of 16 years. 2. Method In their first study Lynn and Vanhanen (2002) published IQs for 81 countries obtained from studies in which intelligence tests have been administered to samples of the populations. In their second study the number of studies in which intelligence tests have been administered to samples of the populations and for which IQs were calculated was increased to 113 countries (Lynn & Vanhanen, 2006). For most of these countries IQs were measured by the Progressive Matrices, a nonverbal reasoning test. In other cases IQs were measured by a variety of other non-verbal tests including the Cattell Culture Fair and the Goodenough Draw-a-Person test. National IQs were calculated in relation to a mean IQ of 100 in Britain and standard deviation of 15. The increases in intelligence known as the Flynn effect were taken into account in these calculations. They also estimated IQs for a further 79 countries for which there is no evidence from intelligence tests. IQs for these 79 countries were estimated from the measured IQs of similar and geographically adjacent countries. For example, the IQ in Afghanistan was estimated at 84 on the basis of the measured IQ of 84 in neighboring Iran and Pakistan. By the use of this method they provided IQs for all the 192 nations of the world with populations over 40,000. Some of the national IQs reported in Lynn and Vanhanen's second study differ slightly from those reported in their first study. The explanation for this is that they had collected more studies in the second study, but the differences are very small. The data for counties' educational achievement in science and mathematics in were taken from the TIMSS 2003 study. This study was carried out by testing representative samples of school students from each country. The student samples were obtained by random sampling. At the first stage of the sampling procedure, 150 public and private schools were selected in each country. At the second stage, random sampling was used to find one class in every school that had two or more classes of students of the required age. The percentage of excluded students was less than five percent from the desired sample. The average number of students tested in a country was 4498 for the grade 4 study and 4777 students for the grade 8 study. The smallest number of Table 2 National IQs and TIMSS scores on math and science Countries IQ Math Science Math Science Armenia Australia Bahrain Belgium Botswana Bulgaria Chile Cyprus Egypt England Estonia Ghana Hong Kong Hungary Indonesia Iran Israel Italy Japan Jordan Korea, Rep. of Latvia Lebanon Lithuania Macedonia Malaysia Moldova Morocco Netherlands New Zealand Norway Palestine Philippines Romania Russia Saudi Arabia Scotland Serbia Singapore Slovakia Slovenia South Africa Sweden Taiwan Tunisia United States 94⁎ 98 83⁎ 99 70⁎ 93 90 91⁎ 81 100 99 71 108 98 87 84 95 102 105 84 106 98⁎ 82 91 91 92 96 84 100 99 100 86 86 94 97 84⁎ 97 89 108 96 96 72 99 105 83 98 Grade 4 449 507 – 544 – – – 511 – 534 – – 565 526 – 395 – 504 566 – – 531 – 529 – – 502 356 536 502 462 – 362 – 527 – 499 – 580 – 486 – – 559 330 521 Grade 4 438 520 – 518 – – – 483 – 538 – – 540 529 – 424 – 517 541 – – 532 – 510 – – 499 306 518 519 470 – 332 – 527 – 502 – 558 – 492 – – 551 317 534 Grade 8 470 506 406 534 369 475 394 459 406 502 531 282 581 526 413 414 492 483 569 428 585 507 432 500 435 505 457 390 535 497 465 395 380 472 505 339 500 473 596 505 492 278 499 580 409 502 Grade 8 455 525 442 515 366 479 421 444 423 540 548 258 550 539 426 458 489 493 546 476 550 512 396 518 451 510 471 399 532 518 496 439 378 472 512 398 510 468 569 517 520 257 522 565 412 526 students tested in a country was 2890 and the largest number was 9829 (Martin, Mullis, Gonzales, & Chrostowski, 2004, pp. 369–375). The students of grade 4 were 10 years old and the students of the grade 8 were 14 years old. The same students were involved in the science and in the mathematics study in both grades. The TIMSS tests included the same tasks in all the participating countries. The tests represented the curricula 118 R. Lynn, J. Mikk / Intelligence 35 (2007) 115–121 Table 3 Correlation coefficients (corrected for attenuation in brackets) between TIMSS 2003 results and IQ Subject Grade All countries Measured IQ countries Mathematics Science Mathematics Science Fourth grade Fourth grade Eighth grade Eighth grade 0.87 (.95) 0.85 (.92) 0.92 (1.0) 0.91 (.99) 0.89 (.97) 0.86 (.93) 0.93 (1.0) 0.91 (.99) in the eighth grade. The reliability of the measures was 0.84 for both grades (Martin et al., 2004, pp. 356–386). The TIMSS 2003 mathematics test was composed and administered analogously to the TIMSS science tests. The mathematics test included items on number, algebra, measurement, geometry, and data. The items were in four cognitive domains: knowing facts and procedures, using concepts, solving routine problems, and reasoning (Mullis, Martin, Gonzales, & Chrostowski, 2004, p.343). About 40% of the items were in multiple-choice format. The international average scale sore was 495 in the fourth grade and 467 in the eighth grade. The reliability was 0.87 for the fourth grade test and 0.89 for the eighth grade mathematics test (Mullis et al., 2004, p.367). In addition, questionnaires were given to students, teachers and school head-teachers. The answers give information about the learning and teaching background in schools and homes. The publication of the 2003 Trends International Mathematics and Science Study (TIMSS) makes it possible to examine the relation between the national scores in mathematics and science and the national IQs calculated by Lynn and Vanhanen in their more recent of the countries as well as possible (Martin et al., 2004, p.359). The science tests included items on life science, chemistry, physics, earth science, and environmental science. The test items were on three cognitive domains — factual knowledge, conceptual understanding, and reasoning and analysis (Martin et al., 2004, p.356). Altogether 152 items in the test for the fourth grade level and 189 test items were used in the test for the eighth grade. The items in the tests were in multiple-choice format. 1200–2000 students answered every test item. Every correct answer was awarded one or two points depending on the complexity of the item. The international average scores were 489 in the fourth grade and 474 Table 4 The correlation coefficients between TIMSS test results in grade 8 science and school characteristics (the labels are explained below of the table) Mean TIMSS Mean TIMSS IQ Lynn Textbook Class size Teach Univ+ Sch CL Low Stud safety Life expectancy >50% EDH GDP USD Computer Study desk Parent Univ IQ Lynn Textbook Class size Teach Univ + Sch Cl Low 1,00 Stud safety Life expectancy >50% EDH GDP USD 1.00 −0.63 −0.74 −0.74 −0.51 1.00 0.83 0.47 0.43 Computer Study desk 1.00 0.91 0.34 − 0.46 0.45 − 0.49 1.00 0.31 − 0.38 0.43 − 0.51 1.00 − 0.15 0.16 − 0.28 1.00 − 0.15 0.19 1.00 − 0.31 0.68 0.72 0.66 0.75 0.39 0.22 − 0.66 − 0.37 0.22 0.45 − 0.31 − 0.65 1.00 0.62 1.00 − 0.78 0.55 0.66 0.73 0.44 − 0.73 0.61 0.70 0.69 0.43 − 0.29 0.01 0.12 0.22 0.27 0.56 − 0.34 − 0.38 − 0.43 − 0.45 − 0.41 0.18 0.25 0.28 0.49 0.54 − 0.42 − 0.59 − 0.46 − 0.37 − 0.58 0.38 0.41 0.58 0.55 −0.57 0.52 0.64 0.52 0.40 1.00 0.64 0.36 1.00 0.50 TIMSS — the arithmetical average of the TIMSS grade 8 science sub-tests. IQ — the national values of IQ from Lynn and Vanhanen (2006). Textbook — percentage of students taught by teachers who report textbooks as primary bases for lessons. Teach Univ — percentage of students whose teachers have university degrees or equivalent. Sch Cl Low — percentage of students whose principals assess school climate as low. Stud safety — percentage of students who assess high their being safe in schools. >50% EDH — the percentage of students in schools with more than 50% economically disadvantaged students according to school principals' reports. GDP USD — Gross domestic product (from The World Factbook, 2005). Computer — percent of students who have computer in the home. Study desk — percent of students who have a study desk/table in the home. Parent Univ — percentage of students whose parents finished university or equivalent or higher. R. Lynn, J. Mikk / Intelligence 35 (2007) 115–121 (Lynn and Vanhanen, 2006) publication. The 2003 International Mathematics and Science Study gives scores for grade 4 school children for 25 countries and for grade 8 school children for 46 countries in mathematics (Mullis et al., 2004) and science (Martin et al., 2004). These reports also give sub-test scores for different areas of mathematics and science. 3. Results The average sub-test scores for both subjects and both grades are given in Table 2, together with the national IQs calculated by Lynn and Vanhanen (2006). Six of the IQs are asterisked to show that they are estimated IQ. The remaining 40 are measured IQs. The data for Belgium are for the Flemish-speaking part of the country; the data for Palestine are for the Palestinian National Authority. The IQ of 86 for Palestine is given in Lynn (2006) and the IQ of 97 for Scotland is given in Lynn (1979). The Pearson correlations between the TIMSS 2003 results and IQ are given in Table 3. The correlations are given for all countries (column 3) and for measured IQ countries (column 4). The correlations for all countries and for measured IQ countries are virtually identical, suggesting that the estimated IQs must be accurate. The correlations are also given corrected for attenuation for all countries and for measured IQ countries. The reliability coefficients for calculating the corrections for attenuation were derived as follows. The reliability of the national IQs is given by Lynn and Vanhanen (2006, p.62) as 0.92, derived as the correlation between pairs of IQs obtained from the same country and based on 71 countries for which there are two IQ measures. The reliability of the national mathematics scores is derived as the correlation between the scores of the grade 4 and the grade 8 children (r = 0.92), and the reliability of the national science scores is derived in the same way (r = 0.92). The correlations corrected for attenuation are given in brackets in Table 3. We have considered whether any other variables could explain the national differences in scores in mathematics and science and examined the possible effects of a number of school and demographic variables including teacher training, school climate, economic development. The data were taken from the TIMSS international Science report for grade eight (Martin et al., 2004). The correlations between these and scores in science for the 8 grade students are shown in Table 4. It will be seen that many of the correlations are high but the correlation between TIMSS science scores and IQ is the highest. Regression analysis for the eighth grade science results revealed only one statistically significant predictor — IQ. Coefficients 119 of partial correlation revealed that no characteristic of schooling had a statistically significant relationship with TIMSS results when IQ was partialed out. 4. Discussion The first question addressed in this paper is whether the validity of the national IQs can be demonstrated by showing that they are highly correlated with national scores on tests of mathematics and science. We have examined the relationships between the national IQs and the results of the four TIMMS studies of international achievement in mathematics and science of grade 4 and grade 8 school students published in 2003. The study has found very high correlations between national IQs and the scores obtained in mathematics and science. These correlations range between .85 and .93; and, corrected for attenuation, between .93 and 1.0. We believe that these high correlations establish beyond reasonable dispute that the national IQs have a high degree of validity. The second question addressed in this paper is how to explain the relationships responsible for the national differences in IQ and attainment on tests of mathematics and science. It has been noted by Luo, Thompson, and Detterman (2003) that there are three theories to explain the correlations between IQs and educational attainment at the level of individuals. These are (1) intelligence is a cause of educational attainment “because g is commonly acknowledged as more pervasive in intellectual tasks, and appears to be more biologically rooted than school achievement” (Jensen, 1998, p.68); (2) intellectual abilities are partly a product of education (Ceci, 1991); (3) intelligence, measured by intelligence tests, and scholastic performance are both partly determined by “basic cognitive processes, which are measured using tasks such as simple reaction time, inspection time, and memory recall and recognition tasks” (Luo et al., 2003, p. 67). These authors produce evidence supporting the last of these theories and conclude that “individual differences in mental speed are a main causal factor underlying the observed correlation between general intelligence and scholastic performance in children between the ages of 6 and 13.” These theories can be applied to explain the crosscountry correlations between IQs and the mathematics and science attainment: (1) we can follow Jensen and regard national differences in intelligence (operationalized by IQs) as virtually entirely responsible for differences in mathematics and science attainment; (2) we can follow Ceci (1991) and posit that differences in education have had an effect on both national IQs and 120 R. Lynn, J. Mikk / Intelligence 35 (2007) 115–121 educational attainment in math and science. There is little doubt that education affects IQs at the level of individuals, as Ceci (1991, p.703) has concluded: “Schooling fosters the development of cognitive processes that underpin performance on most IQ tests”. More recently the positive effect of education on IQ has been confirmed by Whaley, Fox, Deary, and Starr (2005) who have found that the level of education is an independent predictor of intelligence at age 64 after controlling for childhood IQ and other predictors. Further evidence has been provided by Blair, Gamson, Thorne, and Baker (2005). (3) We can follow Luo et al. (2003) and regard national IQs and educational attainment in math and science as measures of the same latent construct that they describe as “basic cognitive processes”. The cross-country correlations between the IQs and mathematics and science attainment scores are so high (at between .93 and 1.0, corrected for attenuation) that this appears a plausible theory. This conclusion is supported by genetic studies that have found that the relationship between academic attainment and IQ is largely due to common genetic influences (Wainwright, Wright, Geffen, Luciano, & Martin, 2005; Kovas, Harlaar, Petrill, & Plomin, 2005). The third objective of the paper has been to explore further the relationships between national IQs and mathematics and science attainment and a number of educational, economic and demographic variables. The correlation matrix showing the inter-relationships between these is given in Table 4. It will be seen that a number of these correlations are moderately high. Educational attainment and IQ are correlated with the use of textbooks as primary bases for lessons (r = .34, .31), the percentage of teachers with university degrees (r = .45, .43), the percentage of students whose principals assess the school climate as low (r = − .49,− .51), the percentage of students who assess their schools as safe (r = .68, .66), the percentage of students in schools with more than 50% economically disadvantaged students according to school principals' reports (r = − .78,− .73), the percentage of students who have a computer in the home (r = .66, .70), the percentage of students who have a study desk or table in the home (r = .73, .69), and the percentage of students whose parents had university degrees or equivalent (r =. 44, .43). There are also substantial correlations between national IQs and mathematics and science attainment and gross national per capita domestic product (r = .55, .61) and life expectancy (r = .72, .75). We suggest that these correlations arise from a complex network of relationships. We propose that the two most fundamental are national IQs and per capita GDP (a measure of per capita income) that are correlated at 0.61. We suggest that these two variables have a reciprocal positive feedback relationship such that each has a causal effect on the other, i.e. national IQs have a positive effect on per capita income and per capita income has a positive effect on national IQs. National IQs have a positive effect on per capita income because a nation whose population has a high IQ can earn more, just as individuals with high IQs can earn more than those with low IQs. Conversely, national per capita income has a positive effect on national IQs because a nation whose population has a high per capita income can spend more on education, nutrition and health care, and these have a beneficial effect on intelligence. The other correlations given in Table 4 appear to arise largely as effects of per capita income on educational variables. For instance, national populations with high per capita incomes can afford to buy their children desks (r = .47) and computers (r = .83), and can afford to have smaller classes (r = −.34). National populations with high per capita incomes have high life expectancy (r = .52) probably largely because they can provide a high standard of nutrition and health care and also because intelligence is associated with longevity, as shown by Whaley and Deary (2001). References Baker, D. B., & Jones, D. P. (1993). Creating gender equality: Crossnational gender stratification and mathematical performance. Sociology of Education, 66, 91−103. Barber, N. (2005). Educational and ecological correlates of IQ: A cross-national investigation. Intelligence, 33, 273−284. Barnett, S. M., & Williams, W. (2004). National intelligence and the emperor's new clothes: A review of IQ and the Wealth of Nations. Contemporary Psychology, 49, 389−396. Beaton, A. E., Mullis, I. V. S., Martin, M. O., Gonzales, E. J., Kelly, D. L., & Smith, T. A. (1996). Mathematics achievement in the middle school years. Chestnut Hill, MA: Boston College, TIMSS International Study Center. Beaton, A. E., Martin, M. O., Mullis, I. V. S., Gonzales, E. J., Smith, T. A., & Kelly, D. L. (1996). Science achievement in the middle school years. Chestnut Hill, MA: Boston College, TIMSS International Study Center. Blair, C., Gamson, D., Thorne, S., & Baker, D. (2005). Rising mean IQ: Cognitive demand of mathematics education for young children, population exposure to formal schooling, and the neurobiology of the prefrontal cortex. Intelligence, 33(93), 106. Ceci, S. J. (1991). How much does schooling influence general intelligence and its cognitive components? Developmental Psychology, 27, 703−722. Comber, L. C., & Keeves, J. (1973). Science achievement in nineteen countries. New York: John Wiley. Deary, I. J. (2004). Intelligence differences. In R. L. Gregory (Ed.), The Oxford companion to the mind Oxford, UK: Oxford University Press. Dickerson, R. E. (2006). Exponential correlation of IQ and the wealth of nations. Intelligence, 34, 292−295. Ervik, A. O. (2003). Review of IQ and the wealth of nations. The Economic Journal, 113, 406−408. R. Lynn, J. Mikk / Intelligence 35 (2007) 115–121 Gagne, F., & St. Pere, F. (2002). When IQ is controlled, does motivation still predict achievement? Intelligence, 30, 71−100. Hunt, E., & Sternberg, R. J. (2006). Sorry, wrong numbers: An analysis of a study of a correlation between skin color and IQ. Intelligence, 34, 121−139. Husen, T. (1967). International achievement in mathematics: A comparison of twelve countries. New York: John Wiley. Jensen, A. R. (1998). The g factor: The science of mental ability. Westport: Praeger Publishers. Jones, G., & Schneider, W. J. (2006). Intelligence, human capital, and economic growth: A Beysian averaging of classical estimates (BACE) approach. Journal of Economic Growth, 11, 71−93. Kovas, Y. N., Harlaar, S. A., Petrill, S. A., & Plomin, R. (2005). Generalist genes and mathematics in 7-year-old twins. Intelligence, 33, 473−489. Lloyd, J., & Barenblatt, L. (1984). Intrinsic intellectuality: Its relation to social class, intelligence, and achievement. Journal of Personality and Social Psychology, 46, 646−654. Lubinski, D., & Humphreys, L. G. (1996). Seeing the forest from the tree: When predicting the behavior or status of groups, correlate means. Psychology, Public Policy and Law, 2, 363−376. Luo, D., Thompson, L. A., & Detterman, D. (2003). The causal factor underlying the correlation between psychometric g and scholastic performance. Intelligence, 31, 67−83. Lynn, R. (1979). The social ecology of intelligence in the British Isles. British Journal of Social and Clinical Psychology, 18, 1−12. Lynn, R. (2006). Race differences in intelligence: An evolutionary analysis. Athens, GA: Washington Summit Books. Lynn, R., & Vanhanen, T. (2002). IQ and the wealth of nations. Westport, CT: Praeger. Lynn, R., & Vanhanen, T. (2006). IQ and global inequality. Athens, GA: Washington Summit Books. Lynn, R., Hampson, S., & Magee, M. (1984). Home background, intelligence, personality and education as predictors of unemployment in young people. Personality and Individual Differences, 5, 549−558. Martin, M. O., Mullis, I. V. S., Gonzales, E. J., & Chrostowski, S. J. (2004). TIMSS 2003 international science report. : IEA, Boston College. Matarazzo, J. D. (1972). Wechsler's measurement and appraisal of intelligence. Baltimore: Williams and Wilkins. 121 McDaniel, & Whetzel (2004). IQ and wealth of nations: Prediction of GDP not dependent on precise g estimates. New Orleans: Conference of the International Society for Intelligence Research. Mullis, I. V. S., Martin, M. O., Gonzales, E. J., & Chrostowski, S. J. (2004). TIMSS 2003 international mathematics report Boston: IEA, Boston College. Templer, D. I., & Arikawa, H. (2006). Temperature, skin color, per capita income, and IQ: An international perspective. Intelligence, 34, 121−139. Tewes, U. (2003). Germany. In J. Georgas, L. G. Weiss, F. van der Vijver, & D. H. Saklofske (Eds.), Culture and children's intelligence Amsterdam: Academic Press. Volken, T. (2003). IQ and the wealth of nations: A critique of Richard Lynn and Tatu Vanhanen's recent book. European Sociological Review, 19, 411−412. Voracek, M. (2004). National intelligence and suicide rate: An ecological study of 85 countries. Personality and Individual Differences, 37, 543−553. Wainwright, M. A., Wright, M. J., Geffen, G. M., Luciano, M., & Martin, N. G. (2005). The genetic basis of academic achievement on the Queensland Core Skill Test and its shared genetic variance with IQ. Behaviour Genetics, 35(2), 133−145. Walberg, H. J. (1984). Improving the productivity of America's schools. Educational Leadership, 41, 19−27. Weede, E., & Kämpf, S. (2002). The impact of intelligence and institutional improvements on economic growth. Kyklos, 55, 361−380. Whaley, L. J., Fox, H. C., Deary, I. J., & Starr, J. M. (2005). Childhood IQ, smoking, and cognitive change from age 11 to 64 years. Addictive Behaviors, 30(1), 77−88. Whaley, L. J., & Deary, I. J. (2001). Longitudinal cohort study of childhood IQ and survival up to age 76. British Medical Journal, 322, 1−5. Whetzel, D.L. and McDaniel, M.A. (in press). Prediction of national wealth. Intelligence. The World Factbook (2005). Retrieved June, 2005 http://www.cia.gov/ cia/publications/factbook/ Yule, W., Gold, R. D., & Busch, C. (1982). Long term predictive validity of the WPPSI: An 11 year follow-up study. Personality and Individual Differences, 3, 65−71.