Richard Lynn has been arguing since 1994 for the so-called developmental theory of sex differences in intelligence. It states that, while males and females have equal IQ up to age 15 or so, eventually men begin to outperform women, resulting in a sex difference of roughly 4 IQ points in adulthood. It should come as no surprise that this view is controversial, but is it scientifically sound?
In this book, Lynn conducts an extensive analysis of sex differences in IQ and cognitive abilities to support his theory. I did find myself disagreeing with his presentation of the evidence at various points throughout, and will highlight some of my criticisms below.
Brain size
In chapter one, Lynn explains the “problem” the developmental theory is intended to solve: In spite of many experts’ insistence that there is no sex difference in average IQ, there remains the fact that a large sex difference in brain size does exist. It has also been established that brain size and IQ are positively correlated. The most recent meta-analysis puts this correlation at either r=0.24 or r=0.29 in healthy samples, depending on which statistical methods are used (Pietschnig et al., 2022). Lynn concludes from this, “it appears to follow that males should have a higher average IQ than females”.
I don’t find this reasoning convincing. The relatively low correlation between IQ and brain size implies that about 90% of the variation in IQ scores is explained by other factors. Male and female brains differ on average after controlling for size, and it seems likely that those remaining differences play a role in IQ as well, possibly compensating for the male advantage in brain volume. Lynn complains that other researchers – those defending what he derisively calls the “equalitarian theory” – never elaborate on these supposed compensatory mechanisms. However, I have seen concrete explanations proposed – see for instance Bouchard (2014) about the possibly greater “local network efficiencies” of female brains.
Nevertheless, Lynn thinks that a sex difference of 4 IQ points should be expected based on the larger male brain size. He arrives at that estimate by multiplying the sex difference in brain size adjusted for body size from Ankney (1992) with the correlation between brain size and IQ. This calculation seems dubious to me. It is based on rather old data, and the estimate of the IQ-brain size correlation used by Lynn is probably a bit too high. Furthermore, Ankney’s finding of a sex difference in relative brain size has been criticized by Forstmeier (2011), who argues that Ankney used inappropriate statistical methods and in his own analysis finds a minor female advantage in relative brain size.
More importantly, though, why should we expect the sex difference in IQ to reflect the male advantage predicted from the sex difference in brain volume? Other group differences in IQ cannot usually be accounted for by variation in brain size alone. For instance, although Europeans average both larger brains and higher IQ scores than Africans, the difference in brain size explains at most a small fraction of the gap in IQ (Wicherts et al., 2010). Similarly, Arctic populations like the Inuit have larger brains but score lower on IQ tests than Europeans, according to earlier work by Lynn himself (Lynn, 2006, p. 153).
Developmental differences
Lynn’s theory predicts that there should be little difference between the sexes in childhood IQ, but that in older subjects a male advantage should increasingly emerge. He collects data showing exactly that in chapter 2, table 2.1. He also critically discusses the findings of Rojahn & Naglieri (2006) who claim that his developmental theory was not supported in their investigation of sex differences on the Naglieri Nonverbal Test. Lynn believes that these authors hid a male IQ advantage among 16 and 17-year-olds by grouping subjects between the ages of 15 and 17 together, thereby balancing the aforementioned difference against a female advantage at age 15. This criticism does not seem accurate to me. After accessing the paper myself, I found that Rojahn & Naglieri separately report sex differences among subjects with mean age 17 (SD = 1), who took the version of the test appropriate for their age, in table 2. They argue that the male advantage in this group (0.7 IQ points) is too small to support Lynn’s view. One might disagree with that reasoning, but they are not the only authors who have produced findings inconsistent with Lynn's prediction of an emerging male advantage (see for instance Iliescu et al., 2016; Keith et al., 2011; Keith et al., 2008; Reynolds et al., 2008).
Interestingly, Lynn actually cites one of the studies mentioned above, Keith et al. (2011), to support his theory, even though those authors themselves comment that their statistical models suggest “no developmentally related Sex differences in g”. This is possible because he focuses not on the sex difference in g, but instead simply averages sex differences in 4 “broad abilities” from table 7 of that paper and reports the result. Given that the paper investigates sex differences in more than just these four abilities, this approach seems obviously incorrect to me, and Lynn does not attempt to justify it.
Raven’s matrices
In chapter 4, Lynn discusses sex differences on Raven’s progressive matrices. He collects 33 adult samples and reports a median male advantage of 0.30d (4.5 IQ points). This is in line with his own previous research, but conflicts with the results of a recent meta-analysis of “Sex differences in inductive reasoning”. While that meta-analysis also found male advantages for the Standard Progressive Matrices (1.65 IQ points based on 37 samples) and the Advanced Progressive Matrices (2.85 IQ points based on 29 samples), they are smaller than Lynn’s estimate for the same tests (Waschl & Burns, 2020).
I would have expected Lynn to comment on this study, given that part of its stated aim was to update the evidence - collected by Lynn - “for a sex difference in the RPM tests”. Unfortunately, he does not. Maybe he thinks that his own research is more representative because, although it is based on fewer studies, most of them use samples of the general population, whereas Waschl & Burns include data on university students as well. They do carry out a moderator analysis, however, showing that sample population did little to affect their results. Also, if anything, I would expect university samples to show larger sex differences in favor of males. This at least would be implied by the greater male variability hypothesis and the fact that more women than men attend university (suggesting that they are less highly selected).
Other tests of general intelligence
In chapter 6, Lynn turns his attention to “other tests of general intelligence” meaning tests other than Raven’s matrices and the Wechsler tests which are reviewed in chapter 5. He has compiled 43 studies showing a median male advantage of 0.23d (3.45 IQ points). His analysis contains some mistakes, however.
For instance, he reports a male advantage of 0.09d (1.35 IQ points) for 30-year-olds on the British Ability Scales, citing Dykiert et al. (2009). But as that paper makes clear, the cognitive data for these adults was actually collected when they were just 10 years old.
Lynn also reports a sex difference of 0.14d (2.1 IQ points) favoring men for US standardizations of the Cognitive Ability Test in a large sample of 50,000 17-year-olds. This too is a mistake as a difference of this magnitude was found only in the standardization sample of the CogAT 7 which included just 3295 17-year-old participants. Sex differences in the same age group for the CogAT 4, 5 and 6 amount to 0.05d (0.75 IQ points), 0.02d (0.3 IQ points) and -0.02d (0.3 IQ points female advantage) respectively. The sample size reaches 50,000 as reported by Lynn only when considering all 17-year-old participants across all 4 CogAT standardizations together (Lakin, 2013).
I was further surprised to see Lynn report a male advantage for the US standardization sample of the Woodcock-Johnson III. My understanding is that Lynn calculated sex differences from Camarata & Woodcock (2006) not for “General Intellectual Ability” but for “Fluid reasoning”, and that he included only participants between the ages of 19 and 79 whilst excluding the college sample. This analysis is never explained or justified, though. Applying it to the standardization data of the original WJ test and the WJ-R would have produced female advantages of 0.5 and 1.7 IQ points. This data is also available in the paper referenced by Lynn, but ignored by him.
Overall, Lynn’s selection of studies - at least in this chapter - seems biased. He includes 6 studies of sex differences assessed with the German IST. These mostly rely on small samples, but report relatively large male advantages. A bigger Romanian standardization sample of the same test that found no sex difference is not included, even though Lynn does report on other Romanian data available from the same source (Iliescu et al., 2016). A Slovak study using the IST that found a female advantage is ignored as well (Kohút et al., 2016). Similarly, Lynn does include three findings of male advantages on the Differential Aptitude Test but excludes the large US standardizations, the latest of which produced a female advantage (Feingold, 1988).
It is certainly easy to claim that of 43 studies “in all but one […] males obtained higher IQs than females” when one deliberately ignores studies with a different result. In general, Lynn could stand to be more transparent with how he selected the data discussed throughout his book. This section is about “other tests of general intelligence” which, I assumed, was meant to be a catch-all category for tests devised by psychologists to measure g. Lynn also includes a sample of the Swedish scholastic aptitude test, however, which like the American SAT is presumably primarily intended to detect “college readiness”. I don’t necessarily object to this inclusion, as similar tests have been found to be highly correlated with IQ. Still, many countries organize some kind of university entrance exam which could on this basis be included in an analysis of sex differences in IQ and some of those do show female advantages (Deary et al., 2007; Zhang & Tsang, 2014).
Sex differences in g
In chapter 8, Lynn investigates sex differences in g. He proposes that the best estimate of the sex difference in g is the mean of five studies using multi-group confirmatory factor analysis, which according to Lynn has become the “preferred method” for measuring differences in g.
My impression is that many researchers still use different methods, although some of them have indeed been criticized. Yet, even accepting Lynn’s argument here, I have two further issues with his conclusions.
First, he estimates the sex difference in g at 2.4 IQ points based on the mean effect size of those aforementioned five studies. One of those studies is van der Sluis et al. (2006), for which Lynn reports a male advantage of 0.3d (4.5 IQ points). Those authors, however, find that the sex differences observed on different indices of the WAIS-III in their study are not attributable to g. I don’t understand where Lynn gets his male advantage from.
Second, Lynn's 2.4 point estimate of the sex difference in g is smaller than his 4 point estimate of the male advantage in IQ. But he thinks that it is the latter which best represents the sex difference in intelligence, and concludes that “the adult male advantage in intelligence is attributable to both g and other abilities”.
I don’t think it makes sense to claim that the sex difference in intelligence is larger than the sex difference in g. After all, g is often interpreted as “general intelligence” with full-scale IQ usually thought to be a somewhat imperfect approximation of g.
Group differences on the g factor should also be less arbitrary than group differences in IQ. Test developers could easily influence IQ by simply adding or removing items and subtests that favor one group over another.
Different intelligence tests all seem to measure the same underlying g however (Johnson et al., 2004). As a result, group differences in g should be less malleable.
Some research also indicates that the predictive validity of IQ is primarily attributable to the variance explained by the g factor (Ree & Carretta, 2022).
For these reasons, it is probably the sex difference in g one should be interested in, not the sex difference in IQ.
Specific Abilities
Chapter 10 and 11 of this book are devoted to sex differences in specific cognitive abilities. I only want to comment on Lynn’s review of verbal ability here. He argues that there is a female verbal advantage in adolescence, but a male advantage in adulthood.
The adult male advantage is supposedly proven by the higher male verbal IQ assessed by the WAIS, the female advantage in adolescence by results from educational surveys and the DAT. The problem is that these tests all measure verbal ability in different ways. The WAIS' verbal comprehension index, for instance, includes the Information subtest, which can be more accurately said to test general knowledge. The DAT, on the other hand, uses tests of spelling and English skills. It makes more sense to compare verbal IQ from the WAIS to verbal IQ from the WISC, but the WISC already shows a male verbal advantage for an age group (6 to 16-year-olds) in which girls outperform boys on other verbal tests. Thus, Lynn's conclusion relies on using different tests to assess verbal ability at different ages. A rather questionable methodology. (Some) other studies of verbal ability do, in fact, show a female advantage in adulthood in contrast to his review (Hyde & Linn, 1988).
Conclusion
Lynn does discuss an impressive amount of data in this book. Nevertheless, his review is let down by his unwillingness to explain his methodological decisions. In some cases, Lynn and the authors he cites seem to come to different conclusions from the same data, and since he rarely shares the specifics of his analysis with the reader, it is difficult to determine the cause of such discrepancies.
I don't know how common an issue this is, as I did not follow up on all of Lynn's references. My criticisms are mostly limited to those instances where I was already familiar with the research discussed and found that his interpretation did not match my own.
Furthermore, it is often unclear by which criteria studies are included in Lynn’s dataset. I certainly got the impression that at least in some sections his selection is biased towards studies consistent with his theory and against those that contradict it.
As for his conclusions, Lynn is definitely right about the existence of sex differences in the development of cognitive abilities. Those are also reported by educational surveys like the American NAEP, though in contrast to Lynn’s IQ data they tend to find not just a (usually slight) increase in the male advantage in math tests as students age but also a growing female advantage in reading (Reilly et al., 2019).
I am less sure about Lynn’s specific estimates for the sex difference in g or on various IQ tests. It is true that most studies on adults report either no sex difference at all or a male advantage, which maybe indicates that a small difference is indeed present. However, I am not convinced by Lynn’s 4 point estimate.
It is also possible that sex differences on IQ tests reflect sampling issues because low IQ subjects in general - among whom men may be overrepresented - and low IQ men in particular are seemingly harder to recruit for IQ testing, especially in adulthood after they have stopped attending school. This argument is made by Dykiert et al. (2009) among others.
References
Ankney, C. D. (1992). Sex differences in relative brain size: The mismeasure of woman, too? Intelligence, 16, 329–336
Bouchard Jr, T. (2014). Genes, Evolution and Intelligence. Behavior Genetics, 44
Camarata, S., & Woodcock, R. (2006). Sex differences in processing speed: Developmental effects in males and females. Intelligence, 34, 231–252
Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence, 35, 13–21
Dykiert, D., Gale, C., & Deary, I. (2009). Are apparent sex differences in mean IQ scores created in part by sample restriction and increased male variance? Intelligence, 37
Feingold, A. (1988). Cognitive gender differences are disappearing. American Psychologist, 43, 95–103
Forstmeier, W. (2011). Women have Relatively Larger Brains than Men: A Comment on the Misuse of General Linear Models in the Study of Sexual Dimorphism. Anatomical Record (Hoboken, N.J. : 2007), 294, 1856–1863
Hyde, J. S., & Linn, M. C. (1988). Gender differences in verbal ability: A meta-analysis. Psychological Bulletin, 104, 53–69
Iliescu, D., Ilie, A., Ispas, D., Dobrean, A., & Clinciu, A. I. (2016). Sex differences in intelligence: A multi-measure approach using nationally representative samples from Romania. Intelligence, 58, 54–61
Johnson, W., Bouchard, T. J., Krueger, R. F., McGue, M., & Gottesman, I. I. (2004). Just one g: Consistent results from three test batteries. Intelligence, 32, 95–107
Keith, T., Reynolds, M., Roberts, L., Winter, A., & Austin, C. (2011). Sex differences in latent cognitive abilities ages 5 to 17: Evidence from the Differential Ability Scales—Second Edition. Intelligence, 39, 389–404
Keith, T. Z., Reynolds, M. R., Patel, P. G., & Ridley, K. P. (2008). Sex differences in latent cognitive abilities ages 6 to 59: Evidence from the Woodcock–Johnson III tests of cognitive abilities. Intelligence, 36, 502–525
Kohút, M., Halama, P., Dočkal, V., & Žitný, P. (2016). Gender differential item functioning in slovak version of intelligence structure test 2000-revised. Studia Psychologica, 58, 238–250
Lakin, J. (2013). Sex differences in reasoning abilities: Surprising evidence that male–female ratios in the tails of the quantitative reasoning distribution have increased.
Lynn, R. (2006). Race differences in intelligence: An evolutionary analysis.
Pietschnig, J., Gerdesmann, D., Zeiler, M., & Voracek, M. (2022). Of differing methods, disputed estimates and discordant interpretations: The meta-analytical multiverse of brain volume and IQ associations. Royal Society Open Science, 9, 211621
Ree, M., & Carretta, T. (2022). Thirty years of research on general and specific abilities: Still not much more than g. Intelligence, 91, 101617
Reilly, D., Neumann, D. L., & Andrews, G. (2019). Gender differences in reading and writing achievement: Evidence from the National Assessment of Educational Progress (NAEP). American Psychologist, 74, 445–458
Reynolds, M. R., Keith, T. Z., Ridley, K. P., & Patel, P. G. (2008). Sex differences in latent general and broad cognitive abilities for children and youth: Evidence from higher-order MG-MACS and MIMIC models. Intelligence, 36, 236–260
Rojahn, J., & Naglieri, J. A. (2006). Developmental gender differences on the Naglieri Nonverbal Ability Test in a nationally normed sample of 5-17 year olds. Intelligence, 34, 253–260
van der Sluis, S., Posthuma, D., Dolan, C. V., de Geus, E. J. C., Colom, R., & Boomsma, D. I. (2006). Sex differences on the Dutch WAIS-III. Intelligence, 34, 273–289
Waschl, N., & Burns, N. (2020). Sex differences in inductive reasoning: A research synthesis using meta-analytic techniques. Personality and Individual Differences, 164, 109959
Wicherts, J. M., Borsboom, D., & Dolan, C. V. (2010). Evolution, brain size, and the national IQ of peoples around 3000 years B.C. Personality and Individual Differences, 48, 104–106
Zhang, S., & Tsang, M. (2014). Gender gap in the National College Entrance Exam performance in China: A case study of a typical Chinese municipality. Asia Pacific Education Review, 16, 27–36
Enjoy fast, free delivery, exclusive deals, and award-winning movies & TV shows with Prime
Try Prime
and start saving today with fast, free delivery
Loading your book clubs
There was a problem loading your book clubs. Please try again.
Not in a club? Learn more
Join or create book clubs
Choose books together
Track your books
Bring your club to Amazon Book Clubs, start a new book club and invite your friends to join, or find a club that’s right for you for free.
Download the free Kindle app and start reading Kindle books instantly on your smartphone, tablet, or computer - no Kindle device required.
Read instantly on your browser with Kindle for Web.
Using your mobile phone camera - scan the code below and download the Kindle app.
Discover the latest buzz-worthy books, from mysteries and romance to humor and nonfiction. Explore more
Frequently bought together
$15.20
Get it as soon as Monday, Mar 11
In Stock
Total price:
To see our price, add these items to your cart.
Choose items to buy together.
Customers who viewed this item also viewed
Page 1 of 1 Start overPage 1 of 1
About the author
Follow authors to get new release updates, plus improved recommendations.
