the negative flynn effect a systematic literature review

• DOI: 10.1016/J.INTELL.2016.10.002
• Corpus ID: 51729528

The negative Flynn Effect: A systematic literature review

• E. Dutton , D. Linden , R. Lynn
• Published 1 November 2016
• Psychology, Economics
• Intelligence

Tables from this paper

110 Citations

No negative flynn effect in france: why variations of intelligence should not be assessed using tests based on cultural knowledge, what causes the anti-flynn effect a data synthesis and analysis of predictors, measurement-invariant fluid anti-flynn effects in population—representative german student samples (2012–2022), a negative flynn effect in kuwait: the same effect as in europe but with seemingly different causes.

• Highly Influenced

Intelligence Differences across Years: A Trend Analysis

The flynn effect in germanophone preschoolers (1996–2018): small effects, erratic directions, and questionable interpretations, is the flynn effect related to migration, a negative flynn effect in khartoum, the sudanese capital, fl ections about intelligence over 40 years, the flynn effect in estimates of premorbid intellectual functioning in an australian sample, 50 references, who discovered the flynn effect a review of early studies of the secular increase of intelligence.

• Highly Influential

What is intelligence? Beyond the Flynn effect

Is the flynn effect on g: a meta-analysis, one century of global iq gains, secular declines in cognitive test scores: a reversal of the flynn effect, a reversal of the flynn effect for spatial perception in german-speaking countries: evidence from a cross-temporal irt-based meta-analysis (1977–2014), efeito flynn: estudo com adultos portugueses, generational iq gains: spanish data, the secular rise in iqs: in estonia, the flynn effect is not a jensen effect., the end of the flynn effect: a study of secular trends in mean intelligence test scores of norwegian conscripts during half a century, related papers.

Showing 1 through 3 of 0 Related Papers

The Negative Flynn Effect: a Systematic Literature Review

Intelligence 59 (2016) 163–169

Contents lists available at ScienceDirect

Intelligence

The negative Flynn Effect : A systematic literature review

Edward Dutton a,⁎, Dimitri van der Linden b,RichardLynnc a Ulster Institute for Social Research, London, UK b Erasmus University Rotterdam, Department of Psychology , Education, and Child Studies, The Netherlands c University of Ulster, London, UK article info abstract

Article history: The Flynn Effect (rising performance on intelligence tests in the general population over time) is now an Received 18 January 2016 established phenomenon in many developed and less developed countries. Recently, evidence has begun to Received in revised form 9 September 2016 amass that the Flynn Effect has gone into reverse; the so-called ‘Negative Flynn Effect.’ In this study, we present Accepted 13 October 2016 a systematic literature review, conducted in order to discover in precisely how many countries this reverse phe- Available online 20 October 2016 nomenon has been uncovered. Using strict criteria regarding quality of the sample and the study, we found nine studies reporting negative Flynn Effects in seven countries. We also discuss several possible explanations for the Keywords: Negative Flynn Effect negative Flynn Effect as an attempt to understand its most probable causes. IQ decline © 2016 Elsevier Inc. All rights reserved. Dysgenics Immigration

1. Introduction (see Flynn, 2012). The effect has been found in Latin America, for exam- ple Brazil (Colom, Flores-Mendoza, & Abad, 2007)andArgentina(Flynn Numerous studies have shown that scores on intelligence tests (i.e., & Rossi-Case, 2012) and in the former Communist Bloc (e.g. Estonia; IQ tests) have substantially increased during much of the twentieth cen- Must, Must, & Raudik, 2003). The average global increase of 3 IQ points tury. This increase in IQ scores was first reported in the United States by per decade has also been confirmed in a recent meta-analysis by Runquist (1936) and was later confirmed by Smith (1942) and Pietschnig and Voracek (2015). Tuddenham (1948). Cattell (1951) reported a similar trend in England Although one criterion of the Flynn Effect is that it applies to the full after it became apparent that the increase in IQ scores was a largely IQ score, it should be noted, that it nevertheless is driven more strongly global phenomenon found in many different countries (summarized by rises on certain kinds of ability. Relative to scores produced by verbal in Lynn, 2013). tests, there have been greater gains in scores produced by nonverbal, This phenomenon has been designated the Flynn Effect after the re- performance-based measures like Raven's Progressive Matrices and view by Flynn (1984, 1987, 2012) who first documented it on the Wechsler performance subtests (Flynn, 2012). These types of tests are Stanford-Binet and Wechsler tests. A Flynn Effect is distinct from simple associated with fluid intelligence, suggesting less of a rise in crystalized sample fluctuations in IQ scores in annually assessed cohorts. More spe- intelligence (which reflects the influence of education, such as vocabu- cifically, the effect is understood to be established if there is an overall lary). A notable exception is the increasing scores produced by the increase in IQ score, rather than merely an increase on one IQ domain Wechsler verbal subtest Similarities, although this subtest relates to el- (e.g., verbal or spatial ability), and if the increase is continuous over ements of reasoning not required by the other subtests comprising the many years or over a relatively long period of time. In this regard, Wechsler Verbal IQ composite. Flynn (1984) found there had been a 13.8 point increase in IQ scores be- The Flynn Effect means that an individual in the most recent cohort tween the years 1932 and 1978, amounting to a rise of 0.3 points per will attain a higher IQ score when set against the norms of an earlier co- year or 3 points per decade. More recently, the Flynn Effect was further hort than he will when set against his own. Accordingly, it means that evidenced by calculations of IQ score gains between 1972 and 2006 on the test will overestimate a subject's IQ by an increasing amount the versions of the Stanford-Binet (SB), Wechsler Adult Intelligence Scale longer the time-span is between the norming of the test and its admin- (WAIS), and Wechsler Intelligence Scale for Children (WISC) (Flynn, istration. This has clear practical implications for any organization that 2009). The Flynn Effect (Flynn, 2012) is not only occurring in developed seeks to use an IQ test for any purpose. Indeed, Flynn (2012) has countries, but it is increasingly being found in developing countries as highlighted the problem of US states having rules whereby those who well, such as Kenya , Turkey , Dominica, Saudi Arabia, China and Sudan are mentally retarded cannot be executed if found guilty of murder. If the test from which their IQ score was gleaned, such as a test at school, ⁎ Corresponding author. was not appropriately normed then it will not provide them with the E-mail address: [email protected] (E. Dutton). correct IQ score and, accordingly, an intellectually impaired person

http://dx.doi.org/10.1016/j.intell.2016.10.002 0160-2896/© 2016 Elsevier Inc. All rights reserved. 164 E. Dutton et al. / Intelligence 59 (2016) 163–169 may be executed. This example, provided by Flynn, however, also im- IQ tests and this restriction had to be adhered to. plies that the increase in IQ scores may not be a genuine increase in in- (4) It had to be an overall decline in IQ score on the test, rather than telligence. This notion is in line with more recent studies that have simply a decline restricted to one of its domains. Again, this is an suggested that increases in IQ scores may not necessarily reflect an in- essential dimension to the Flynn Effect itself and must therefore crease in the latent intelligence factor, g (te Nijenhuis, van Vianen, & be employed in any analysis of the reversal of this effect. van der Flier, 2007; te Nijenhuis & Van der Flier, 2013). All in all, the (5) It had to be original research. Articles which cited other articles Flynn Effect is an important phenomenon in the study of intelligence were ruled out. and beyond. Recently, however, a number of prominent studies have reported a reversal in the Flynn Effect, beginning in the latter decades of the twen- Based on these criteria we searched for the literature using the fol- tieth century. A number of these studies (e.g. Sundet, Barlaug, & lowing Prisma procedure. We first conducted the literature search Torjussen, 2004) draw upon annually accrued conscript military data, using Google Scholar. Google Scholar was chosen because of its expan- which involves the vast majority of the nation's males in any given an- sive nature . Unlike many academic search engines, it is not restricted nual cohort. Accordingly, they allow us to conclude that there is a con- to peer-reviewed journals, let alone high-impact journals. Thus, the tinuous decline in IQ scores over time and that this is a real use of this database would be more likely to allow us to uncover under- phenomenon and not simply a blip. graduate and postgraduate theses, book content, and working papers. More recently, a negative Flynn Effect has also been reported in a We searched for papers up until 2016. number of other countries, such as France (Dutton & Lynn, 2015). In ad- The literature search strategy involved selecting the appropriate key dition, a series of studies have noted declines in proxies for intelligence, words. We identified these from the titles and abstracts of already pub- such as in the lengthening of reaction times between 1900 and 2000 lished literature on the Flynn Effect and through brainstorming synony- (Woodley, te Nijenhuis, & Murphy, 2014), a decline in colour discrimi- mous or roughly synonymous terms. nation ability (Woodley, te Nijenhuis, & Murphy, 2015) and a rise in The next step was the ‘Data Management’ of the Prisma procedure. fluctuating facial asymmetry (Woodley & Fernandes, 2016), something In line with the Data management stage, all titles were copied into a which has been shown to be weakly negatively correlated with intelli- Word Document, numbered, and marked with whether they were to gence. These sets of findings by Woodley and colleagues seem to indi- be accepted or rejected and, if so, on what grounds. cate that approximately a century ago, a decline in the latent factor of Finally, we conducted the Prisma Selection Process: Based on the genuine intelligence began, which until now has largely been cloaked general criteria outlined above and on further analyses of the content by an increase in scores on IQ tests. Yet, with the increase in IQ scores of the articles (e.g. title, abstract, text), the three authors decided collec- slowing down (see Pietschnig & Voracek, 2015) or even going into re- tively, based on the above-described criteria, which papers were to be verse, the decline in g may start to become apparent on IQ tests. included and there had to be complete agreement for this to be the case. Woodley et al. (2015) term this the ‘co-occurrence model’ which states In line with Prisma guidelines, we now outline our Literature Search that the Flynn Effect occurs on the more environmentally influenced Process: The term ‘Anti-Flynn Effect’ yielded 139 articles or theses with and less g-loaded aspects of intelligence, while dysgenics concomitantly this term in the text. Of these, seven articles were original articles that occurs on the more g-loaded and more heritable dimensions. had found an overall negative Flynn Effect. However, this was reduced If there truly is a trend toward declining intelligence as indicated by to 6 because Ronnlund, Carlstedt, Blomstedt, Nilsson, and Weinehall's a negative Flynn Effect then this would be a phenomenon with poten- (2013) negative Flynn Effect on Swedish conscript data between 1992 tially serious implications. This is because there is evidence that national and 1993 was only over a year. IQ is associated with economic prosperity (Lynn & Vanhanen, 2012), po- The term ‘Negative Flynn Effect’ yielded 37 hits and 3 articles within litical stability (Lynn & Vanhanen, 2012)andscientificachievement our inclusion criteria, including one which the previous term had not (Rindermann & Thompson, 2011). Therefore in the present study, we uncovered. We rejected Pietschnig and Gittler's (2015) German data be- wish to establish in precisely which countries there has been a reported cause the test it drew upon, namely the 3DC, is mainly a test of spatial negative Flynn Effect. We aim to do this by conducting an exhaustive intelligence. and systematic literature search, as we will outline below. We then ‘Reverse Flynn Effect’ yielded 13 hits but none within our inclusion wish to establish its most likely causes and to address any possible criteria. anomalies. In order to establish this, we conducted a systematic litera- The term ‘Reversal of the Flynn Effect’ uncovered 129 hits of which 8 ture review the details of which we will now present. unique hits were within our inclusion criteria. Among those excluded was a study in Portugal (Schelini, Almeida, Duarte, & Primi, 2011) 2. Method which found a negative Flynn Effect over five years among recruits to the Portuguese gendarmes. However, not only was this not a represen- 2.1. Literature search tative sample, but the two samples employed had significantly different sex compositions and sizes. This also uncovered Bandeira, Costa, and In conducting this review, we followed the PRISMA guidelines Arteche (2012) who found, overall, no generational gain in IQ in a (http://www.prisma-statement.org/). We decided to use the following study of children aged between 6 and 12 on both the Ravens and eligibility criteria for including studies in the review: Draw-a-Man. Their data came from Porto Alegre in Brazil, which they argued was an elite area. But, again, this was not within our inclusion (1) The study related to a population sample, or a sample likely to be criteria. representative of the country's population. The term ‘Decline in IQ’ yielded 1890 hits of which 2 were within our (2) A negative Flynn Effect was noted over a considerable period of inclusion criteria. Using alternative words such as ‘decline of IQ’ or ‘de- time, which was set at 5 years. It is appreciated that this border cline of the IQ’ did not lead to unique and/or relevant hits above and be- involves some degree of arbitrariness. This amount of time was yond the first alternative. selected, however, because, from the studies of which we were The term ‘Fall in intelligence’ unearthed 58 hits, of which none were already aware showing a negative Flynn Effect, there had never relevant. Among those excluded were Cancian and Klein (2015) who been a rise in scores after 5 years of decline. found that the IQ score of US marine officer recruits had fallen by 2.6 (3) The effect had to be on an IQ test, rather than on some proxy for points between 1980 and 2014. Clearly, this is not representative of this such as a scholastic test. This is very important, such that we the US population. ‘Fall of intelligence,’ (77 hits) ‘fall of the intelligence’ were comparing like with like. The Flynn Effect was observed on (26 hits) and ‘fall in the intelligence’ (11 hits) did not reveal additional E. Dutton et al. / Intelligence 59 (2016) 163–169 165 unique hits. The term ‘End of the Flynn Effect’ unearthed 252 hits of section. Lynn (2011) has calculated a dysgenic fertility ratio for which 7 were relevant and one (Sundet et al., 2004) a new relevant hit. most of the countries in which we have observed a negative Reading the relevant articles revealed a further potential article in Flynn Effect. This was based on the difference in fertility between the form of Cotton et al. (2005). However, this found zero gains on the more and less educated men and women. Unfortunately, these Coloured Progressive Matrices in Australia when comparing 1975 and data are rather old. The Norwegian data was from the period 2003. Although this may imply that there has been a negative Flynn Ef- 1925–1930 while the Dutch data was from 1920 to 1925. With fect, this does not constitute one. A small negative Flynn effect was all the other countries, it was from 1966 to 1967. Nevertheless, found among the 8, 10 and 11 year olds but not among the 7, 9, and despite these limitations regarding the age of the data, they po- 10 year olds. We also found, using this method, Koivunen (2007), tentially imply that dysgenics may have begun earlier or which reported a negative Flynn Effect in Finland up to 2001. proceeded more speedily in some countries than others and Parallel to the search in Google Scholar, we conducted the same that this may have led to later differences in the extent and search using Scopus. Due to the more limited nature of Scopus' sample, speed of the negative Flynn Effect. More specifically, the dysgenic one highly relevant finding, in the form of a thesis, was absent. The only fertility that may have caused a possible decline in genetic intel- new potentially germane hit was Flynn (2012) Are We Getting Smarter? ligence may not have been directly accompanied by a decline in This revealed that on the CPM, among British 14.5–15.5 year olds be- IQ scores, as the latter would have also been prone to more envi- tween 1979 and 2008 IQ had decreased by 1.88 points. But when com- ronmental factors that have caused an increase in IQ. As such, it is paring 7.5–15.5 year olds over the same period it had increased by 4.26 useful to test what association they have with more recent nega- points. tive Flynn Effects. We also consulted the most recent meta-analysis of the Flynn Effect in general (Pietschnig & Voracek, 2015). This meta-analysis found most, but not all, of the relevant hits we had found. This meta-analysis, in it- self, revealed a large number of Flynn Effect reversals or cessations, 3. Results but they were either at the domain level only or on unrepresentative samples. For example, the Brazilian study was based on a very specific Based on the literature search detailed in the Method section, and sample from a wealthy area of Brazil. The excluded studies were Brazil within our inclusion criteria, we found reports of a negative Flynn Effect (2), Canada (1), Germany (2), Netherlands (2), South Africa (1), Spain in seven countries. The results can be seen in Table 1. The nine articles (1), Sudan (1), UK (1), and USA (22). Although these studies did not included, draw upon the following tests: reach the criteria for being included in the present review, in order to be complete, in the supplementary material, we have added informa- (1) Sundet et al. (2004) used the General Ability Test, an IQ test de- tion about those studies that showed declines in subgroups or decline veloped by the Norwegian army in 1954. It is composed of in specific cognitive abilities (see Appendix 1). Having found these re- Words, Numbers and Shapes and conscripts are given a GA (Gen- sults, which we will report below, we also planned to obtain some pre- eral Ability) score, which corresponds to an IQ score. liminary information about the possible causes of the Negative Flynn (2) Woodley and Meisenberg's (2013) meta-analysis of tests of Effect. The literature reveals a number of possible causes that have Dutch adults used the GATB = General Aptitude Test Battery. been proposed. These are: This measures 9 different ‘aptitudes’ among which are verbal ap- titude, numerical aptitude and spatial aptitude. (1) Immigration. It has been consistently found that non-Western (3) Teasdale and Owen's (2008) study drew upon the Borge Prien's ethnic minorities in Western countries, such as Sub-Saharan Af- Prove, which is an IQ test used by the Danish army on recruits ricans and South Asians, on average, score lower on IQ tests since 1961. It is comprised of logical, verbal, numerical and spa- than Europeans (Lynn, 2006). We tested the possible influence tial reasoning tests. of immigration on negative Flynn effects by correlating the (4) Shayer and Ginsburg's (2007, 2009) studies drew upon the Pia- mean per capita immigration level between 1981 and 2015 getian test: An IQ test developed for children. Piaget's theory fo- with each country's magnitude of negative Flynn Effect. We cuses on interviewing the subjects to discover why they took these data from Migration Policy (2016), which presents answered in a particular way. mean per capita immigration for each country at 5 year intervals (5) Dutton and Lynn's (2013) study drew upon annual average re- from 1950 to 2015. sults of the Finnish Peruskoe, which literally translates as ‘Basic (2) Sex. Several studies indicated that there are sex differences in in- test.’ This is an IQ test developed by the Finnish army composed telligence and that, in adult samples, males have a slight advan- of Numbers, Words and Shapes tests. These results were report- tage of somewhere between 3 and 5 IQ-points (≈0.3 SD) over ed in Koivunen (2007) up to 2001, a thesis which was sent to females (Lynn & Irwing, 2004). In as far as this is indeed the them by a Finnish army researcher, as well as in correspondence case, a difference in sex ratio may be accompanied with a change with the same Finnish army researcher for 2008–9. in average IQ score. (6) Korgesaar's (2013) Estonian study drew upon the Raven Stan- (3) Parental age. Older age of fatherhood has been associated with dard Progressive Matrices (SPM) test, which is a widely-accepted reduced offspring intelligence, possibly caused by decreasing test of general intelligence and, as such, the study is within our sperm quality (Saha et al., 2009). It has also been shown that in- inclusion criteria. creasing maternal age is negatively associated with IQ (e.g. (7) (7) Dutton and Lynn (2015) drew upon the French WAIS Myrskylä, Silventoinen, Tynelius, & Rasmussen, 2013) and that (Wechsler Adult Intelligence Test) IV manual. maternal age is increasing at the same rate as paternal age in Western countries (e.g. Wehner, Kambskard, & Abrahamson, 2002). Data on age at first motherhood by country was available From Table 1 it can be seen that in the majority of the studies the de- from the CIA World Factbook for 2013 (CIA World Factbook, cline ranges between 0.38 and 4.30 IQ points per decade. The Estonian 2016). study seems to be somewhat of an outlier with a decline of 8.4 IQ points (4) Dysgenics. Lynn (2011) has argued that the negative association per decade. Taking the un-weighted average of all the studies, the mean between intelligence and fertility in Western countries, especial- decline per decade in the studies would be 3.18 points. When excluding ly among women, should eventually be reflected in declining IQ the rather high value of Estonia, the average decline in the remaining scores, for reasons we will explore more in our Discussion seven studies becomes 2.44 IQ points per decade. 166 E. Dutton et al. / Intelligence 59 (2016) 163–169

Table 1 Negative Flynn Effect per country.

Country Age Test Years Type IQ (decline per Reference decade)

Norway 18–19 General ability 1996–2002 All conscripts in every year 0.38 Sundet al., 2004 Borge Priene's Denmark 18–19 1998–2003/4 All conscripts in every year 2.70 Teasdale & Owen, 2008 Prove 10,023 over 5 cohorts: 1975, 2000, 2001, 2002, 2003 (each cohort Britain 11–12 Piagetian 1975/2003 4.30 Shayer & Ginsburg, 2007 roughly equal in size) Britain 13–14 Piagetian 1976/2006 2006: N 446, 2007: N 357 (total: 793) 2.50 Shayer & Ginsburg, 2009 Woodley & Meisenberg, Netherlands Adults GATB 1975/2005 Meta-analysis 1.35 2013 Dutton & Lynn, 2013; Finland 18–19 Peruskoe 1998–2009 All conscripts 1998–2001 and 2008/9 2.0 Koivunen, 2007 France Adults WAIS III & IV 1999/2008–9 Two representative groups of 79 3.8 Dutton & Lynn, 2015 Estonia 18–19 Raven SPM 2001/2005/2012 Representative student sample: 2001: 573, 2005: 417, 2012: 338 8.4 Korgesaar, 2013

3.1. Possible explanations (Migration Policy, 2016). When including all countries in the review, the correlation was virtually zero (r(7) = 0.033, p = 0.94). However, As part of our analysis, we decided to discuss and, where possible, to it has to be noted that Estonia did not only appear to be an outlier re- test a number of possible causes of the negative Flynn Effect. These were garding the IQ decline, but also regarding immigration, because it was as follows: the only country that showed negative immigration numbers. When conducting the calculations again, but this time excluding Estonia, the correlation became r(6) = 0.802 and reached marginal significance 3.1.1. Sex ratio (p = 0.055). Nevertheless, we have already noted that, based on the One of the possible explanations for a change in IQ scores is the sam- percentage of immigrants in the most reliable samples, immigration is ple sex ratio. This notion is based on the literature suggesting that there unlikely to have a large influence. Accordingly, this association may be may be sex differences in average IQ. For example, Lynn and Irwing underpinned by a factor, which underlies both dysgenics and high im- (2004) have suggested that in adults, women score, on average 3 to 5 migration levels, such as degree of societal development or putative IQ points lower than men. If this is indeed the case then a change in amount of time since industrialization. That is to say, highly developed the sex ratio of samples may accompany a change in IQ. However, in countries may, because they are highly developed, have high levels of the studies in the present review, sex ratio is unlikely to have caused immigration. the effects found. First, the Danish and Finnish samples consist of males only. In addition, the samples that do have mixed sex are very 3.1.3. Dysgenics similar over time regarding sex ratios. For example, the Estonian 2001 A third possibility that has been suggested is a so-called process of sample is 43% male while the 2012 sample is 42% male. For the follow- dysgenics on intelligence (e.g. Dutton & Lynn, 2015)whichhasbeenar- ing samples: - Shayer and Ginsburg (2007, 2009), Woodley and gued to be occurring due to the negative association between intelli- Meisenberg (2013), and Dutton and Lynn (2015) - the sex ratio could gence and fertility in modern populations (Lynn, 2011). Higher not be reliably retrieved. However, given the information described fertility of less intelligent parents had already been noted anecdotally above, it can be concluded that sex ratio is unlikely to be a determining in Galton (1869). Based on this, one would expect that IQ would already factor in the change in IQ scores that we report in this review. have shown a relatively large decline since the beginning of the 20th century, especially at its heritability has been estimated at around 0.8 3.1.2. Immigration (Lynn, 2011). This, however, clearly did not happen as the Flynn Effect A number of studies have found that immigrants from developing showed that IQ in fact went up for almost a century. countries have average IQs that are significantly lower than the Europe- One possible explanation for this is that a possible dysgenic effect an average (e.g. Lynn, 2006). Therefore, ideally, we would be able to cor- may be obscured by a strong increase in environmental quality that relate the immigration figures per capita and per year for each country has boosted phenotypic IQ. Specifically, there is some agreement that where the negative Flynn Effect has been found with the extent of IQ de- the Flynn Effect may not fully reflect a genuine rise in intelligence, be- cline. However, this would be very complicated because immigrants cause it occurs on the less g-loaded parts of the tests (Lynn, 2011; te from different countries will have different IQs, will go on to attain dif- Nijenhuis et al., 2007; te Nijenhuis & Van der Flier, 2013). Rather, it ferential socioeconomic statuses, and potentially be reacted to different- may reflect a rise in pattern spotting ability (which is correlated with in- ly by hosts due to a variety of factors, including religion . Focussing on telligence), something precipitated, possibly, by living in more educated the Nordic data, Sundet (correspondence quoted in Dutton, 2014, society, as argued by Flynn (2012). He argues that modern society p.243) has noted that: “Men from (South) Asian and African countries makes us increasingly look at the world through ‘scientific spectacles’ have around 5–6 IQ points lower than non-immigrants. Yet, they and, accordingly, examine it analytically, boosting performance IQ, es- seem to comprise no more than around 2–3% of the conscripts in this pecially on similarities. As this is ultimately underpinned by intelli- period. If there would be any effect then this would deflate the total gence, it would have a genotypic limit and if genotypic intelligence mean IQ by around a maximum of 0.1–0.2 IQ points.” As such, it simply were declining then the imperfect nature of the IQ test as a measure cannot fully explain the decline. Also, conscript data from Finland is par- of intelligence would mask this, but only up to the genotypic limit. ticularly important in assessing the immigration hypothesis. Finland did Once this limit was reached, any genotypic decline in IQ would become not experience any significant third world immigration until around visible on the IQ tests. As mentioned above, this is known as the Co-oc- 1992 (see Dutton & Lynn, 2013). However, the conscripts in 1997 currence Model. It is possible that this is what has happened because would have been mainly born in 1978 when the non-white population dysgenic fertility - a negative association between intelligence and of Finland was vanishingly small. numbers of children - has been observed in Denmark, Sweden, Finland, Despite the limitations outlined above and the small N (=7) for and a number of other countries reviewed in Lynn (2011). country level, we decided to calculate the correlation between the IQ Indeed, if Flynn's ‘scientific spectacles’ explanation is accurate then decline per decade and average immigration between 1950 and 2015 we would expect to see, prior to an overall negative Flynn Effect, a E. Dutton et al. / Intelligence 59 (2016) 163–169 167 negative effect on verbal and mathematical IQ concomitant with a pos- found corroborative evidence in other proxies for IQ, such as colour dis- itive effect on other parts of the test. This is, indeed, what we see in the crimination (Woodley & Fernandes, 2016), meaning that the simplest studies we excluded. Khaleefa, Sulman, and Lynn (2009) found that Su- explanation is that these studies are tapping into the same phenome- danese Full-scale IQ increased 2.05 points per decade between 1987 and non. Moreover, Woodley et al. (2015) have responded to the most re- 2007, but Verbal IQ decreased by 1.65 points over the period. Colom, cent criticisms with further evidence for their case. It might also be Andres-Pueyo, and Juan-Espinosa (1998) reported a decline in Spanish countered that their evidence would predict IQ losses throughout the verbal reasoning (male and female −0.3) and mathematical reasoning twentieth century, rather than beginning at the end of it. But, as already (male −2.4; female −2.1) between 1979 and 1995 but a rise on ab- discussed, the argument is that these would be being covered-up by stract reasoning (and also Ravens) sufficient to create an overall Flynn massive rises on a particular intelligence sub-ability that was partly re- Effect. lated to g. Eventually, the ceiling of this ability would be reached and the Besides such differential effects on subtests, we would also expect to losses would start to reveal themselves on IQ tests. In addition, Woodley see a slowing down of the Flynn Effect before it ultimately ceased, be- and Fernandes (2016) have shown that the Flynn Effect is not primarily cause the Flynn Effect itself would be partly g-loaded (with g in decline) occurring on general and heritable intelligence factor g, whereas the and there would be a limit to the extent to which the environment can negative Flynn Effect does seem to occur on g. The negative Flynn Effect raise IQ scores. The meta-analysis of the Flynn Effect by Pietschnig and displays a Jensen Effect, and is mainly occurring on the more heritable Voracek (2015) does indeed show that IQ gains since the 1980s had abilities (Woodley of Menie & Dunkel, 2015). We would suggest that considerably slowed down. The gains were also increasingly non-linear these different lines of evidence favour Flynn's argument that the in this period. Flynn Effect is an environmental phenomenon. The more educated and science-focused modern world forces us to adopt scientific specta- 3.1.4. Maternal age cles causing us to reach our phenotypic limit on certain intelligence abil- Finally, we tested maternal age, which could also be regarded as a ities, even though g is declining due to dysgenic fertility. This shows up proxy for age at which people become parents more generally. Unfortu- as a secular rise in IQ scores, cloaking the decline in g, but this has nately, we could not find data, by country, on average age of first father- reached its limit, so the decline in g is now showing up on the IQ tests hood by year, only averages for different years, rendering the datasets as a negative Flynn Effect. incomparable. We did, however, find the average maternal age per Fourth, and finally, we discussed maternal age. We tested the possi- country (N = 7) to negatively correlate to the level of IQ decline bility that this may play a role in intelligence decline, but the direct cor- (r = −0.88). Despite the small sample size, the correlation reached sig- relation between IQ decline and the country's average maternal age was nificance (p b 0.01). Nevertheless, as the correlation was negative this negative (−0.88). Thus, at this stage, maternal age is not persuasive as seems to indicate that among the countries showing a negative Flynn an argument for the IQ decline. Effect, the ones with the highest average maternal age show smaller de- In the present study, we must emphasize that even though we iden- clines in IQ. As such, it seems unlikely that maternal age is a factor in IQ tified several studies showing a decline in IQ, there is currently still a decline. much larger pool of studies showing an increase in IQ. For example, Flynn (2012) has reported a positive Flynn Effect in the USA among 4. Discussion adults between 1995 and 2001 of 3.06 points per decade, when compar- ing the WISC III and the WISC IV. Among US children, between 1989 and In the present study we conducted an exhaustive literature search in 2001, he reports a rise of 3.36 points per decade based on the same tests. order to find studies reporting a negative Flynn Effect. Based on a set of Moreover, in their recent meta-analysis, Pietschnig and Voracek (2015) strict criteria which we detailed in the Method section, we found a se- confirmed the 3 points per decade average increase in IQ. Therefore, it is ries of articles reporting a decline in general IQ scores over a period of imperative to discuss our negative Flynn Effects in the light of the posi- at least five years. Notably, the most useful studies were those from Nor- tive Flynn Effects findings in the literature. way, Denmark and Finland because they involve the vast majority of A first possible explanation is that the studies we report are simply males in each annual cohort and provided clear evidence of a year-on- outliers in a much larger set of longitudinal studies on IQ. However, year negative Flynn Effect. given that some of the studies we reported drew upon very large and pop- Building on the results of this systematic literature search, we ulation-representative samples, this explanation seems rather unlikely. discussed four possible causes of the negative Flynn Effect. First, we A second possibility is that a phenotypic increase in IQ may have took into account sex ratio. We concluded that this sampling issue is largely overshadowed a possible genotypic decline in IQ and that the very unlikely to play a role because the negative Flynn Effect can be latter is only more recently starting to show up in several representative found in samples of almost the entire male cohorts in certain countries datasets. If we are correct that the positive Flynn Effect is primarily an and where we have a mixed-sex sample the sex-balance of the cohorts environmental effect then we should not be especially surprised that is roughly the same. it operates differently in different parts of the world. For example, if The second possibility we considered was immigration. We argued the positive Flynn Effect would be partly the result of an increase in that in some samples (e.g., Finland, Dutton & Lynn, 2013; Koivunen, the ‘scientific worldview’ of the population, then the Flynn Effect may 2007) immigration could not entirely have caused the decline in IQ vary in line with the level that a country has reached in adopting such scores because the level of immigrants in those cases was too small to a view. Specifically, factors that may influence the extent to which the have any significant effect on average IQ. population adopts this ‘scientific-spectacles’ view, such as education, The third option we considered was dysgenics. Unfortunately, we wealth, and, possibly, religion, differ between countries. could not extract dysgenic data for all the countries in which we Regarding the trade-offs between forces that drive Flynn Effects and found a negative Flynn Effect. Therefore, we could not reliably analyse forces that drive negative Flynn Effects, it may not be a coincidence that the relationship with IQ decline. Nevertheless, the notion that dysgenics the countries in which we found negative Flynn Effects are some of the may play a role has been suggested in several other studies. For example wealthiest countries with the best distribution of wealth indices. It is it would be in line with Woodley et al. (2014) who have presented ev- possible that because of these optimizing factors, those countries are idence that reaction times have declined across the twentieth century in among the first to reach the boundaries of the environmental effects a number of Western countries and they show that reaction times ro- on IQ. Consequently, they may also be among the first to indicate a de- bustly correlate with intelligence. Woodley et al.'s (2014) studies on re- cline in IQ scores. Future research may want to elaborate on these pos- action times have been heavily criticised (e.g. Woods, Wyma, Yund, et sibilities and focus more strongly on disentangling the various al., 2015). However, it can be countered that Woodley and others have explanations for Flynn versus negative Flynn Effects. 168 E. Dutton et al. / Intelligence 59 (2016) 163–169

5. Limitations Supplementary data to this article can be found online at doi:10. 1016/j.intell.2016.10.002. There are a number of limitations to this analysis. Ideally, in order to establish and understand the causes of the negative Flynn Effect, we Acknowledgements need large samples, annual cohorts over a longer period of time, and, where possible, the ability to rule out potential confounding factors We would like to thank Jakob Pietschnig, Jim Flynn, and an anony- such as sex and immigration. We only have this for Finland and, to a mous reviewer for their useful comments and suggestions on previous lesser extent, Denmark and Norway. Here, we have, from the military versions of this article. conscript data, samples which are almost the entire male population of a certain age, year-on-year. The impact of immigration is likely to be very small, as we have seen, and this is especially so in the case of Fin- References land. Ideally, we would need more samples of this quality and also fe- male samples of this quality as it is possible that the negative Flynn Bandeira, D. R., Costa, A., & Arteche, A. (2012). The Flynn Effect in Brazil: Examining gen- erational changes in the draw-a-person and in the Raven's colored progressive matri- Effect is working at a different rate among females. ces. Revista Latinoamericana de Psicologia, 44,9–18. A second limitation can be found in our ability to test hypotheses as Cancian, M., & Klein, M. (2015). Military officer quality in the all-volunteer force. National to the possible causes of the negative Flynn Effect. As discussed, some of Bureau of Economic Research (Working Paper 21372). Cattell, R. B. (1951). The fate of national intelligence: Test of a thirteen year prediction. these data have had to be drawn upon because they are the best that we Eugenics Review, 17,136–148. have. Also, because a negative Flynn Effect has only been found in a CIA World Factbook (2016). https://www.cia.gov/library/publications/download/ small number of countries, we are left with a limitation of power, leav- download-2013/index.html fi fi fi Colom, R., Andres-Pueyo, A., & Juan-Espinosa, M. (1998). Generational IQ gains: Spanish ing it dif cult to have con dence in any correlational nding in this data. Personality and Individual Differences, 25,927–935. study. In some cases, such as the immigration correlation, it raises the Colom, R., Flores-Mendoza, C., & Abad, F. (2007). Generational changes on the draw-a- question of whether the correlation reflects immigration causing a man test: A comparison of Brazilian urban and rural children tested in 1930, 2002 and 2004. Journal of Biosocial Science, 39,79–89. country to have a negative Flynn Effect or whether countries that are Cotton, S., Kiely, P., Crewther, D., Thompson, B., Laycock, R., & Crewther, S. (2005). Anor- more developed, and so have a negative Flynn Effect, are more likely mative and reliability study of the Raven's coloured matrices for primary school aged to have high levels of immigration. children from Victoria, Australia. Personality and Individual Differences, 39,647–659. Dutton, E. (2014). Religion and intelligence: An evolutionary analysis. London: Ulster Insti- tute for Social Research. 6. Concluding statements Dutton, E., & Lynn, R. (2013). A negative Flynn Effect in Finland, 1997–2009. Intelligence, 41,817–820. – Although, the Flynn Effect is a well-known phenomenon that has Dutton, E., & Lynn, R. (2015). A negative Flynn Effect in France, 1999 2008/9. Intelligence, 51,67–70. been replicated in many samples and in various countries, several theo- Flynn, J. R. (1984). The mean IQ of Americans: Massive gains 1932 to 1978. Psychological ries predict a negative Flynn Effect (e.g., Lynn, 2011; Woodley, te Bulletin, 95,29–51. Nijenhuis, & Murphy, 2013). In the current study, we identified the Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101,171–191. high quality samples reported in the literature that have reported Flynn, J. (2009). What is intelligence? Beyond the Flynn Effect. New York: Cambridge Uni- such a negative Flynn Effect. Even though identifying these studies versity Press. may be a crucial step in this area, future research should further scruti- Flynn, J. R. (2012). Are we getting smarter? Rising IQ in the twenty first century. Cambridge: Cambridge University Press. nize the evidence in light of other studies that continue to report an in- Flynn, J. R., & Rossi-Case, L. (2012). IQ gains in Argentina between 1964 and 1998. crease in IQ scores. For example, it is unclear why some Western Intelligence, 40,145–150. countries, such as the USA, continue to display a positive Flynn Effect. Galton, F. (1869). Hereditary genius. London: MacMillan. Khaleefa, O., Sulman, A., & Lynn, R. (2009). An increase of intelligence in Sudan, 1987– With regard to some of the theories we discussed, it seems to us that 2007. Journal of Biosocial Science, 41,279–283. the simplest explanation for the negative Flynn Effect has already been Koivunen, S. (2007). Suomalaismiesten kognitiivisen kykyprofiilin muutokset 1988–2001: presented by scholars such as Woodley et al. (2013). The Industrial Rev- Flynnin efektiä suomalaisessa aineistossa? (Pro Gradu) Jyväskylä University. olution had two key and related effects on IQ. Firstly, it set off massive Korgesaar, M. (2013). Flynni Efekti Esinemine Eesti Abiturientide Seas Raveni Testi Pohjal. (Seminaritöö) Tartu Ulikool. gains on IQ scores by establishing an environment which compelled Lynn, R. (2006). Race differences in intelligence: An evolutionary analysis. Augusta, GA: us to think in more scientific way, compelled us to become more educat- Washington Summit Publishers. ed, and saturated us with knowledge, information, and novel problems. Lynn, R. (2011). Dysgenics: Genetic deterioration in modern populations. London: Ulster In- stitute for Social Research. This process would have pushed certain latent intelligence abilities to Lynn, R. (2013). Who discovered the Flynn Effect? A review of early studies of the secular their limit and, if this occurred at a sufficient pace, then the imperfect increase of intelligence. Intelligence, 41,765–769. nature of the IQ test means that it would overwhelm whatever else Lynn, R., & Irwing, P. (2004). Sex differences on the progressive matrices: A meta-analysis. Intelligence, 32,481–498. was happening and show up as an IQ rise. However, the Industrial Rev- Lynn, R., & Vanhanen, T. (2012). Intelligence: A unifying construct for the social sciences. olution concomitantly set off a quite different process. It led to the de- London: Ulster Institute for Social Research. velopment of reliable contraception and low child mortality, meaning Migration Policy (2016). http://www.migrationpolicy.org/programs/data-hub/ international-migration-statistics people only needed to have very few children and the more intelligent Must, O., Must, A., & Raudik, V. (2003). The secular rise in IQs: In Estonia the Flynn Effect is would be more able to achieve this. Accordingly, the previous positive not a Jensen Effect. Intelligence, 31,461–471. correlation between intelligence and fertility became negative (Lynn, Myrskylä, M., Silventoinen, K., Tynelius, P., & Rasmussen, F. (2013). Is later better or worse? Association of advanced parental age with offspring cognitive ability among 2011). Moreover, this positive correlation was a so-called Jensen Effect: half a million young Swedish men. American Journal of Epidemiology, 177,649–655. it was occurring on general intelligence and on the more heritable di- Pietschnig, J., & Gittler, G. (2015). A reversal of the Flynn effect for spatial perception in mensions of intelligence. The Flynn Effect would disguise this on IQ German-speaking countries: Evidence from a cross-temporal IRT-based meta-analy- – – tests but once the ceiling of the Flynn Effect was reached, we would sis (1977 2014). Intelligence, 53,145 153. Pietschnig, J., & Voracek, M. (2015). One century of global IQ gains: A formal meta-anal- see it slowing down and then going into reverse as a negative Flynn Ef- ysis of the Flynn Effect (1909–2013). Perspective on Psychological Science, 10, fect. There would, perhaps, be two dimensions to this. One would be 282–306. purely genetic and the other would be environmental: as the society be- Rindermann, H., & Thompson, J. (2011). Cognitive capitalism : The effect of cognitive abil- ity on wealth, as mediated through scientific achievement and economic freedom. comes less intelligent the degree to which it innovates an increasingly Psychological Science, 22,754. cognitively stimulating environment would, set off the by the Industrial Ronnlund, M., Carlstedt, B., Blomstedt, Y., Nilsson, L., & Weinehall, L. (2013). Secular Revolution, slow down, meaning it would outpace the decline g to a trends in cognitive test performance: Swedish conscript data 1970–1993. Intelligence, 41,19–24. lesser and lesser extent until it stopped outpacing it altogether, mani- Runquist, E. A. (1936). Intelligence test scores and school marks in 1928 and 1933. School festing itself as a negative Flynn Effect. & Society, 43,301–304. E. Dutton et al. / Intelligence 59 (2016) 163–169 169

Saha, S., Barnett, A. G., Foldi, C., Burne, T. H., Eyles, D. W., et al. (2009). Advanced paternal Wehner, C., Kambskard, M., & Abrahamson, P. (2002). Demography of the family: The case age is associated with impaired neurocognitive outcomes during infancy and child- of Denmark. UK: University of York. hood. PLoS Medicine, 6, e1000040. http://dx.doi.org/10.1371/journal.pmed.1000040. Woodley, M., & Fernandes, H. (2016). Showing their true colors: Secular declines and a Schelini, P., Almeida, L., Duarte, C., & Primi, R. (2011). Efeito Flynn: Estudo com adultos Jensen Effect on color acuity – More evidence for the weaker variant of Spearman's Portuegesos. Psicologia Educação e Cultura, 15,407–415. other hypothesis. Personality and Individual Differences, 88,280–284. Shayer, M., & Ginsburg, D. (2007). Thirty years on - a large anti-Flynn Effect? The Piaget- Woodley, M., & Meisenberg, G. (2013). In the Netherlands the anti-Flynn Effect is a Jensen ian test volume & heaviness norms 1975–2003. Journal of Educational Psychology, 77, Effect. Personality and Individual Differences, 54,871–876. 25–41. Woodley of Menie, M. A., & Dunkel, C. (2015). In France, are secular IQ losses biologically Shayer, M., & Ginsburg, D. (2009). Thirty years on—A large anti-Flynn Effect? 11: 13-year caused? A comment on Dutton and Lynn. Intelligence, 53,81–85. http://dx.doi.org/10. olds. Piagetian tests of formal operations norms 1976–2006/7. British Journal of 1016/j.intell.2015.08.009. Educational Psychology, 79,409–418. Woodley, M., te Nijenhuis, J., & Murphy, R. (2013). Were the Victorians cleverer than us? Smith, S. (1942). Language and non-verbal test performance of racial groups in Honolulu The decline in general intelligence from a meta-analysis of the slowing of simple re- before and after a fourteen year interval. Journal of General Psychology, 26,51–93. action time. Intelligence, 41,843–850. Sundet, J. M., Barlaug, D. G., & Torjussen, T. M. (2004). The end of the Flynn effect? A study Woodley, M., te Nijenhuis, J., & Murphy, R. (2014). Is there a dysgenic secular trend to- of secular trends in mean intelligence test scores of Norwegian conscripts during half wards slowing simple reaction time? Responding to a quartet of critical commentar- acentury.Intelligence, 32,349–362. ies. Intelligence, 46,131–147. te Nijenhuis, J., & Van der Flier, H. (2013). Is the Flynn Effect on g? A meta-analysis. Woodley, M., te Nijenhuis, J., & Murphy, R. (2015). The Victorians were still faster than us. Intelligence, 41,802–807. Commentary: Factors influencing the latency of simple reaction time. Frontiers in te Nijenhuis, J., van Vianen, A. E. M., & van der Flier, H. (2007). Score gains on g-loaded Human Neuroscience, 9,452. tests: No g. Intelligence, 35,283–300. Woods, D., Wyma, J., Yund, E., et al. (2015). Factors influencing the latency of simple re- Teasdale, T. W., & Owen, D. R. (2008). Secular declines in cognitive test scores: A reversal action time. Frontiers in Human Neuroscience, 9,31.http://dx.doi.org/10.3389/fnhum. of the Flynn effect. Intelligence, 36,121–126. 2015.00131. Tuddenham, R. D. (1948). Soldier intelligence in World Wars I and II. American Psychologist, 3,54–56.

• Erasmus Research Institute of Management (ERIM)
• Erasmus School of Economics (ESE)
• Rotterdam School of Management (RSM)
• Erasmus MC: University Medical Center Rotterdam
• Erasmus School of Health Policy & Management (ESHPM)
• Erasmus School of Law
• Public Administration (FSS)
• Netherlands Institute for Government (NIG)
• Institute for Housing and Urban Development Studies (IHS)
• Erasmus School of History, Culture and Communication (ESHCC)
• Erasmus School of Social and Behavioural Sciences(ESSB)
• Erasmus School of Philosophy (ESPhil)
• International Institute of Social Studies (ISS)
• Submitting Publications
• Department of Psychology /
• Intelligence (Kidlington) /
• Search: Search

E. Dutton (Edward) , D. van der Linden (Dimitri) and Lynn, R. (Richard)

The negative Flynn Effect: A systematic literature review

Publication.

Intelligence (Kidlington) , Volume 59 p. 163- 169

The Flynn Effect (rising performance on intelligence tests in the general population over time) is now an established phenomenon in many developed and less developed countries. Recently, evidence has begun to amass that the Flynn Effect has gone into reverse; the so-called ‘Negative Flynn Effect.’ In this study, we present a systematic literature review, conducted in order to discover in precisely how many countries this reverse phenomenon has been uncovered. Using strict criteria regarding quality of the sample and the study, we found nine studies reporting negative Flynn Effects in seven countries. We also discuss several possible explanations for the negative Flynn Effect as an attempt to understand its most probable causes.

• View at Publisher
• View at Scopus

Add Content

• Help & FAQ

The negative Flynn effect: A systematic literature review

• Work and Organizational Psychology

Research output : Contribution to journal › Article › Academic › peer-review

Research programs

Access to document.

• 10.1016/j.intell.2016.10.002

T1 - The negative Flynn effect: A systematic literature review

AU - van der Dutton, E

AU - van der Linden, Dimitri

AU - Lynn, R

U2 - 10.1016/j.intell.2016.10.002

DO - 10.1016/j.intell.2016.10.002

M3 - Article

SN - 0160-2896

JO - Intelligence

JF - Intelligence

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• PMC10816863

Measurement-Invariant Fluid Anti-Flynn Effects in Population—Representative German Student Samples (2012–2022)

Sandra oberleiter.

1 Department of Development and Educational Psychology, Faculty of Psychology, University of Vienna, 1010 Vienna, Austria [email protected] (J.P.)

Sabine Patzl

2 International Student Assessment (ZIB), TUM School of Social Sciences and Technology, Technical University of Munich, 80333 Munich, Germany

Jonathan Fries

Jennifer diedrich, martin voracek.

3 Department of Cognition, Emotion, and Methods in Psychology, Faculty of Psychology, University of Vienna, 1010 Vienna, Austria; [email protected]

Jakob Pietschnig

Associated data.

The data presented in this study are not publicly available due to privacy restrictions.

Generational IQ test scores in the general population were observed to increase over time (i.e., the Flynn effect) across most of the 1900s. However, according to more recent reports, Flynn effect patterns have seemingly become less consistent. So far, most available evidence on this phenomenon has been categorized by drawing on the classic fluid vs. crystallized intelligence taxonomy. However, recent evidence suggests that subdomain-specific trajectories of IQ change may well be more complex. Here, we present evidence for cross-temporal changes in measurement-invariant figural reasoning tasks in three large-scale, population-representative samples of German secondary school students (total N = 19,474). Analyses revealed a consistent pattern of significant and meaningful declines in performance from 2012 to 2022. Results indicate a decrease in figural reasoning of 4.68 to 5.17 IQ points per decade (corresponding to small-to-medium effects, Cohen d s from 0.34 to 0.38). These findings may be interpreted as tentative evidence for a decreasing strength of the positive manifold of intelligence as a potential cause of the increasing number of recent reports about inconsistent IQ change trajectories.

1. Introduction

In 1968, generational IQ test score changes were, for the first time, interpreted as genuine cohort effects ( Schaie and Strother 1968 ). These changes were subsequently systematically documented by James Flynn, whose name has since become eponymous for cognitive performance increases in the general population ( Flynn 1984 ). Generational changes were observed to be positive over most of the 20th century, with an estimated IQ increase of three points per decade, and to be differentiated according to intelligence domains, with larger fluid than crystallized IQ gains. On a global level, these increases were observed to amount to about 30, 35, and 25 IQ points from 1909 to 2013 regarding full-scale, fluid, and crystallized intelligence, respectively ( Pietschnig and Voracek 2015 ). Notably, these global changes appeared to be nonlinear, with some periods of strong gains being interspersed with some less substantial ones, but with all changes on balance remaining positive (i.e., all in all indicating IQ gains rather than losses).

These patterns have been broadly consistent across most of the 1900s, although the strength of gains appears to be differentiated according to countries. However, more recent observations have shown that Flynn effect patterns have seemingly become less consistent, showing a deceleration (e.g., USA: Rindermann and Thompson 2013 ), a stagnation (e.g., Australia: Cotton et al. 2005 ), or even a reversal (e.g., Denmark: Dutton et al. 2016 ) of the Flynn effect across different countries.

It has been hypothesized that these unexpected patterns may result from the more fine-grained assessment of cognitive abilities in modern psychometric tests, which provides a more detailed account of domain-specific ability change. Specifically, most of the available evidence about the Flynn effect has primarily been contextualized within the classic taxonomy of Cattell’s differentiation between fluid and crystallized IQ ( Pietschnig et al. 2023 ). However, according to the presently most widely accepted conceptualization of human intelligence (namely, the Cattell–Horn–Carroll model [CHC]; Schneider and McGrew 2018 ), fluid and crystallized intelligence are understood as broad abilities that exist on the same level of abstraction as eight further cognitive domains, all of which are superordinate to several lower-order subordinate abilities.

Within the framework of the CHC model, Flynn effects for specific cognitive abilities have recently been shown to be differentiated in terms of stratum II and, arguably, stratum I CHC abilities ( Lazaridis et al. 2022 ). Intriguingly, stratum II domains either showed (i) positive Flynn effects (e.g., comprehension knowledge, learning efficiency), (ii) negative Flynn effects (e.g., spatial orientation, working memory capacity), (iii) ambiguous trends (fluid reasoning, reaction and decision speed, quantitative knowledge, and visual processing), or (iv) no change (processing speed, reading and writing).

This evidence does not necessarily suggest that subdomain differentiation represents a recent phenomenon, but that it instead might be due to the increasing use of more refined intelligence tests beyond the mere assessment of fluid vs. crystallized IQ and psychometric g in more recent decades. Despite a predominant rise in IQ test and subtest scores over time, the available evidence suggests a negative association of the Flynn effect with psychometric g ( Must et al. 2003 ; Woodley et al. 2014 ; Pietschnig and Voracek 2015 ; for contrasting findings, see Colom and Flores-Mendoza 2001 ). A first direct assessment of the association of g with test score changes supports this idea, showing tentative evidence for cross-temporal decreases in achievement g , which may be a necessary consequence of differing population IQ (sub-)domain trajectories ( Pietschnig et al. 2023 ).

In the traditional approach by Cattell, prior related research has demonstrated that fluid IQ typically showed more significant and more robust gains over time than crystallized IQ ( Flynn 1984 ; Pietschnig and Voracek 2015 ).

However, recent findings of domain-specific changes according to the CHC model indicate ambiguous Flynn effects for fluid/reasoning-related subdomains. For example, while some results regarding matrices tests suggest only trivial effects ( d range = −0.002 to −0.05; Lazaridis et al. 2022 ) or the stagnation of IQ gains ( Colom et al. 2023 ), others indicate a reversal of the moderate Flynn effect over time (1996–2001: d = 0.23; 2001–2008: d = −0.11; Pietschnig et al. 2021 ). Thus, it may be reasonable to assume that fluid intelligence change trajectories are rooted in a more fine-grained assessment of specific subdomains.

A significant challenge when assessing the meaningfulness of the Flynn effect revolves around determining whether changes in test scores reflect actual changes in the population’s ability or merely represent manifestations of differential item functioning across different assessment years (DIF). DIF refers to the phenomenon where the discrepancy in average performance between samples results from variations in the difficulty of items or their ability to differentiate between levels of ability rather than differences in actual abilities as societal norms and cultural understandings evolve. This leads individuals to approach these tests with different levels of knowledge, consequently affecting the perceived difficulty of specific items ( Gonthier and Gregoire 2022 ). Therefore, test score changes can only be meaningfully interpreted as population ability changes rather than a measurement artifact when cross-temporal measurement invariance is established (i.e., meaning that there is no DIF, and item properties have not changed over time; ( Lazaridis et al. 2022 )). In the light of recent evidence for unexpected, ambiguous Flynn effect patterns, such as domain-specific and/or country-specific patterns of stagnation or reversal, some researchers have argued that the Flynn effect may genuinely change its direction overall (e.g., Dutton et al. 2016 ). However, whether these patterns would not be better explained by item drift or domain specificity is still being determined.

To contribute to the examination of the Flynn effects in fluid intelligence, we utilized the figural-reasoning subtest of a widely used Germanophone intelligence test battery (Berliner Test zur Erfassung fluider und kristalliner Intelligenz; BEFKI; Wilhelm et al. 2014 ). The data were collected in 2012, 2018, and 2022 as population-representative samples, totaling about 20,000 German secondary school students.

2. Materials and Methods

Before accessing any data, we preregistered the study design, the analysis plan, and the specific main study hypotheses on the Open Science Framework (OSF; https://osf.io/nd7qr , accessed on 27 December 2023). The analysis code is available at https://osf.io/f96mj/files/osfstorage , accessed on 27 December 2023.

In all, data from 19,474 secondary school students from Germany were available. Sociodemographic sample characteristics are provided in Table 1 .

Sample characteristics according to cohort.

2.1. Berliner Test zur Erfassung Fluider und Kristalliner Intelligenz (BEFKI)

For this study, we examined data from the Berlin Test for the Assessment of Fluid and Crystallized Intelligence (BEFKI; Wilhelm et al. 2014 ), a theoretically grounded intelligence test for secondary school students. It allows for the examination of students in grades 8 through 10, irrespective of the school type they are enrolled in.

The BEFKI has been developed based on the CHC model and comprises two subscales to assess crystallized and fluid intelligence. The fluid intelligence scale consists of three subscales assessing verbal, numerical, and figural task performance. We used data from a parallel form of the figural reasoning subscale for the present study. The psychometric properties of this subscale have been shown to be satisfactory, yielding reliabilities of 0.87 (McDonald’s ω) and concurrent validities of >0.90 with fluid intelligence estimates from the cognitive ability test, a well-established German intelligence test ( Heller and Perleth 2000 ), and associations with listening, orthography, reading, and writing test scores ranging from r = 0.65 to 0.69 ( Wilhelm et al. 2014 ).

2.2. Procedure

Within the formal assessments of the Programme for International Student Assessment (PISA), data from three population-representative cohorts of 15-year-olds were collected in Germany in 2012 (in paper–pencil format), 2018, and 2022 (computer-based administration in these subsequent cohorts).

The 16-item figural reasoning subtest had to be completed in 14 min. Across this item set, respondents were required to recognize and apply the logical rules necessary to identify two missing geometric elements required to complete a sequence of three given geometric figures. Respondents had to select the correct elements out of three potential response alternatives for the respective missing elements. Items were scored as correct when both elements were identified correctly.

2.3. Statistical Analysis

Two approaches were pursued to investigate (measurement-invariant) changes in figural reasoning performance. First, we calculated all pairwise standardized mean differences (Cohen d ) between the raw scores of the 2012, 2018, and 2022 cohorts. Second, we utilized measurement invariance analyses and latent means-based calculations derived from these to quantify IQ test score changes. This latter approach allowed us to disentangle genuine cognitive ability changes from those merely caused by item drift (e.g., due to changes in item difficulty or test administration format; see ( Lazaridis et al. 2022 )). Consequently, we conducted multi-group confirmatory factor analysis (MGCFA) to gradually establish measurement invariance levels from configural to strict invariance across all three cohorts.

Because the figural reasoning subtest yields dichotomous data (responses are scored as correct or incorrect), we assessed configural invariance by constraining thresholds and factor loadings of the latent construct to be equal across groups ( Wu and Estabrook 2016 ). Strict invariance was assessed by additionally constraining residual variances to be equal. Model fit was examined based on comparative fit indexes (CFIs). More restrictive models were adopted when between-cohort CFI changes did not exceed 0.01 ( Cheung and Rensvold 2002 ). Subsequently, we estimated latent means and calculated standardized latent change scores across cohorts.

Effect sizes were calculated to indicate the strength of fluid intelligence changes over time, with positive (vs. negative) values representing performance increases (vs. decreases) over the respective interval (i.e., positive vs. negative Flynn effects). Effect sizes were interpreted according to the well-established thresholds introduced by Cohen, being sorted into small, moderate, or large effects (i.e., absolute d s = 0.2, 0.5, and 0.8, respectively; Cohen 1988 ). Cohen d values of raw and latent scores were transformed into the IQ metric and IQ changes per decade (DIQ) via the following formula: DIQ (interval) = [( d × 15)/interval] × 10 (see Lazaridis et al. 2022 ). Further, we performed between-cohorts analyses of covariance (ANCOVAs), with respondent sex as a covariate, to assess the potential sex-specificity of the Flynn effect.

All analyses were conducted in R 4.0.2 ( R Core Team 2022 ) and RStudio 2022.07.2+576 (R Studio Team 2022), and measurement invariance analyses were performed with the lavaan R package ( Rosseel 2012 ).

Our analyses revealed consistent declines in figural reasoning performance over the observed timespan. Measurement invariance analyses showed the good model fit of strict models compared to the configural model (see Table 2 ), thus suggesting that the BEFKI figural reasoning subscale can be assumed to be fully measurement-invariant across all three (i.e., 2012, 2018, and 2022) cohorts. Therefore, the observed changes can be interpreted as genuine ability changes rather than DIF (e.g., due to changes in test administration format).

Model fit across cohorts.

Note. df = degrees of freedom; CFI = comparative fit index.

Standardized test score changes, determined based on raw scores as well as on latent means (see Figure 1 ), showed consistently significant decreases from 2012 to 2022 (with small-to-medium effect sizes, ranging from d s = −0.38 to −0.34; p s < 0.001; see Table 3 and Figure 2 ). These changes amount to a non-trivial loss estimate of 4.68 to 5.17 IQ points per decade.

Raw (red) and latent (blue) mean test score changes over the three cohorts.

Cohen d and DIQ changes between data collection years with 95% confidence intervals.

Raw score- and latent mean-based between-cohort changes, expressed as Cohen d and DIQ-values.

Note. The bottom left triangular matrix represents latent mean-based changes, the top right triangular matrix raw score-based changes, and table entries in parentheses are estimated DIQ-values (IQ change per decade). Negative values indicate performance declines over time. ** p < .01; *** p < .001.

However, an examination of incremental changes between measurement points showed that the changes appeared to be nonlinear. In the interval between 2012 and 2018, we observed significant decreases in test performance in figural reasoning ( d = −0.33 and −0.25 for raw scores and latent means, respectively; p s < 0.001), representing decreases of about 5.4 to 7.0 IQ points over these six years. Results from the subsequent interval (2018 to 2022) were consistent in terms of effect direction and nominal significance, although only trivial in terms of effect size ( d = −0.05 and −0.09 for raw scores and latent means, with p = < .001 and .01, respectively), corresponding to decreases of 1.5 to 2.8 IQ point over these five years.

Analyses of covariance revealed no statistically significant difference in the observed Flynn effect between boys and girls for any cohort (time by sex p s = 0.126 and 0.166 for raw and latent scores, respectively; see Table 4 ).

Model fits of ANOVAs and ANCOVAs based on raw (latent) score calculations.

Note. df = degrees of freedom; parenthetical values refer to latent changes.

4. Discussion

Here, we investigated evidence for cross-temporal changes in a measurement-invariant figural reasoning task based on population-representative samples of German secondary school students. Our analyses revealed a reversed (i.e., negative) Flynn effect consistent across all cohorts, although these changes appeared to be nonlinear in terms of effect strength. These findings are interesting because figural reasoning represents a fluid intelligence domain which, on the contrary, typically has been observed to yield the most substantial (positive) Flynn effects over time (for a meta-analysis, see Pietschnig and Voracek 2015 ).

These findings provide tentative evidence that the recently emerging, rather conflicting, findings about the Flynn effect may be due to the relatively coarse assessments of cognitive performance that have usually been reported in the pertinent literature (see Pietschnig et al. 2023 ). It could be assumed that more fine-grained assessments (i.e., in terms of CHC-stratum I domains) will beneficially contribute towards clarifying the nature, causes, and meaning of the Flynn effect, as discussed below.

We show non-trivial, measurement-invariant decreases in figural reasoning, which is a central domain of fluid cognitive task performance. This contrasts the global pattern of fluid IQ test scores changes over most of the 1900s ( Pietschnig and Voracek 2015 ). However, recent studies have shown evidence for (partly measurement-invariant) Flynn effect reversals in this very domain in several countries (Austria: Lazaridis et al. 2022 ; Norway: Bratsberg and Rogeberg 2018 ; USA: Dworak et al. 2023 ).

These observations may not solely be attributed to an actual decline in fluid abilities. Instead, studies covering more recent timespans may have investigated test score changes based on more refined intelligence models. They might, therefore, have yielded change scores for more specific cognitive (sub)domains. It thus may be speculated that the past practice of examining IQ test score changes based on distinguishing the rather crude domains of fluid vs. crystallized (and fullscale) IQ sensu Cattell ( Cattell 1957 ) may well have been suboptimal and could inadvertently have masked domain-specific trajectories.

Alternatively, the presently observed unexpected declines may result from a generally reversing Flynn effect globally. In particular, the decreasing strength of the global Flynn effect emerging during the 1980s ( Pietschnig and Voracek 2015 ) has been suggested to be a harbinger of an impending stagnation or even reversal of test score gains. Findings from spatial ability performance changes in Germanophones in recent decades are consistent with this interpretation ( Pietschnig and Gittler 2015 ). However, ambiguous patterns of change within countries and stratum II domains ( Lazaridis et al. 2022 ) suggest a more complex mechanism.

Specifically, it has been argued that changes in ability patterns may result from increased ability differentiation ( Pietschnig et al. 2023 ). According to this idea, one would assume that specific (as opposed to all) abilities are becoming more substantially developed because of the increased specialization of modern-day individuals due to changes in environmental reinforcement. Because g is a statistical consequence of the well-established positive manifold of intelligence ( Spearman 1904 ), the ability gain in some specific domains would lead to a weakening of the intercorrelations among IQ subdomains. This, in turn, would explain the previously observed g -based decreases ( Pietschnig and Voracek 2015 ; Pietschnig et al. 2023 ).

However, a decrease in figural reasoning over time cannot be sufficiently explained by ability differentiation because, in its most salient form, ability differentiation would be expected to lead to increases in each subdomain. In contrast, full-scale IQ and the strength of the positive manifold would be expected to decrease. Instead, it may be speculated that ability changes in specific domains may result from changes in environmental demands. Figural reasoning abilities may have become less relevant for success in modern-day environments.

Conceivably, the increasing use of modern technological devices, such as smartphones, tablets, and computers, could have led to individuals (including school students) spending less time on activities that promote figural reasoning (e.g., reading maps, solving puzzles, or drawing; of note, other researchers have argued for the beneficial effects of technology on population IQ developments, see ( Neisser 1997 )). This would support the gist of previous models postulating IQ changes over time due to social multiplier effects in our ever-changing modern environments ( Dickens and Flynn 2001 ). In this vein, expertise in individual areas is increasingly reinforced through environmental channels, leaving room for genetically based propensities that may promote specialization in a given direction.

Akin to the present results, recent studies have also reported negative Flynn effects in specific domains, such as spatial orientation or working memory capacity ( Lazaridis et al. 2022 ). These findings conform to our observations and may likely be due to a similar mechanism. Modern environments, on the one hand, may reinforce the development of more specific, instead of rather general, ability profiles (but, on the other hand, may no longer reward proficiency in particular specific abilities now seen as obsolete or less expedient). It seems plausible that declines in specific abilities indeed occur. Decreasing task performance in specific domains, such as figural reasoning, could be commensurate with the more general idea of varying and IQ (sub)domain-specific change trajectories, manifesting themselves as differentiated patterns of gain vs. stagnation vs. loss, as evidenced by Lazaridis et al. ( 2022 ), that ultimately may lead to a decrease in the strength of the positive manifold of intelligence.

Strengths and Limitations

The strengths of the current study include the psychometrically unidimensional, measurement-invariant test instrument, the large-scale evidence, the population-representative nature of the samples, and the up-to-datedness of the data. Study limitations to be recognized mainly pertain to several generalizability issues whose relevance is currently unknown: the evidence stems from just one (Western) country, the age range of the test-takers is narrow, the instrument represents a single IQ domain, and amidst the observation period a major technological innovation push—with potential relevance for the topic scrutinized here—took place (in the course of the 2010s, smartphones became ubiquitous).

5. Conclusions

In the present study, we show evidence for a negative Flynn effect in figural reasoning on a one-dimensional, measurement-invariant test. These results may indicate that the increasingly inconsistent patterns of the Flynn effect, as witnessed in a growing number of recent reports, may be a consequence of overly broad assessments of cognitive abilities in the datasets typically available for this line of inquiry. It can be speculated that (sub)domain-specific change trajectories are a consequence of changing environmental demands, leading to a decrease in cognitive ability intercorrelations and a weakening of the positive manifold of intelligence.

Funding Statement

Open Access Funding by the University of Vienna.

Author Contributions

Conceptualization, S.O., S.P. and J.P.; methodology, S.O. and J.P.; software, S.O.; validation, J.F. and J.P.; formal analysis, S.O.; investigation, S.O. and J.P.; resources, J.D.; data curation, S.O. and S.P.; writing—original draft preparation, S.O. and J.P.; writing—review and editing, J.F., S.P., J.D. and M.V.; visualization, J.F.; supervision, M.V. and J.P.; project administration, S.O. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Ethical review and approval were not required because the present study is based on archival data which were collected within the Program of International Student Assessment.

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

Conflicts of interest.

The authors declare no conflicts of interest.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

• Bratsberg Bernt, Rogeberg Ole. Flynn effect and its reversal are both environmentally caused. Proceedings of the National Academy of Sciences of the USA. 2018; 115 :6674–78. doi: 10.1073/pnas.1718793115. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Cattell Raymond Bernard. Personality and Motivation: Structure and Measurement. World Book; Chicago: 1957. [ Google Scholar ]
• Cheung Gordon W., Rensvold Roger B. Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling. 2002; 9 :233–55. doi: 10.1207/S15328007SEM0902_5. [ CrossRef ] [ Google Scholar ]
• Cohen Jacob. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Erlbaum; Mahwah: 1988. [ Google Scholar ]
• Colom Roberto, Flores-Mendoza Carmen. Intelligence and working memory: Relationship between factor g , cognitive complexity and processing capacity. Psicologia: Teoria e Pesquisa. 2001; 17 :37–47. doi: 10.1590/S0102-37722001000100007. [ CrossRef ] [ Google Scholar ]
• Colom Roberto, García Luis F., Shih Pei Chun, Abad Francisco J. Generational intelligence tests score changes in Spain: Are we asking the right question? Intelligence. 2023; 99 :101772. doi: 10.1016/j.intell.2023.101772. [ CrossRef ] [ Google Scholar ]
• Cotton Sue M., Kiely Patricia M., Crewther David P., Thomson Brenda, Laycock Robin, Crewther Sheila G. A normative and reliability study for the Raven’s Coloured Progressive Matrices for primary school aged children from Victoria, Australia. Personality and Individual Differences. 2005; 39 :647–59. doi: 10.1016/j.paid.2005.02.015. [ CrossRef ] [ Google Scholar ]
• Dickens William T., Flynn James R. Heritability estimates versus large environmental effects: The IQ paradox resolved. Psychological Review. 2001; 108 :346–69. doi: 10.1037/0033-295X.108.2.346. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Dutton Edward, Linden Dimitri van der, Lynn Richard. The negative Flynn Effect: A systematic literature review. Intelligence. 2016; 59 :163–69. doi: 10.1016/j.intell.2016.10.002. [ CrossRef ] [ Google Scholar ]
• Dworak Elizabeth M., Revelle William, Condon David M. Looking for Flynn effects in a recent online US adult sample: Examining shifts within the SAPA Project. Intelligence. 2023; 98 :101734. doi: 10.1016/j.intell.2023.101734. [ CrossRef ] [ Google Scholar ]
• Flynn James R. The mean IQ of Americans: Massive gains 1932 to 1978. Psychological Bulletin. 1984; 95 :29–51. doi: 10.1037/0033-2909.95.1.29. [ CrossRef ] [ Google Scholar ]
• Gonthier Corentin, Gregoire Jacques. Flynn effects are biased by differential item functioning over time: A test using overlapping items in Wechsler scales. Intelligence. 2022; 95 :101688. doi: 10.1016/j.intell.2022.101688. [ CrossRef ] [ Google Scholar ]
• Heller Kurt A., Perleth Christoph. Kognitiver Fähigkeitstest für 4. bis 12. Klassen, Revision: KFT 4-12+ R [Cognitive Ability Test for Grades 4 to 12, Revised: KFT 4-12+ R] Beltz-Test; Weinheim: 2000. [ Google Scholar ]
• Lazaridis Alexandros, Vetter Marco, Pietschnig Jakob. Domain-specificity of Flynn effects in the CHC-model: Stratum II test score changes in Germanophone samples (1996–2018) Intelligence. 2022; 95 :101707. doi: 10.1016/j.intell.2022.101707. [ CrossRef ] [ Google Scholar ]
• Must Olev, Must Aasa, Raudik Vilve. The secular rise in IQs: In Estonia, the Flynn effect is not a Jensen effect. Intelligence. 2003; 31 :461–71. doi: 10.1016/S0160-2896(03)00013-8. [ CrossRef ] [ Google Scholar ]
• Neisser Ulric. Rising scores on intelligence tests: Test scores are certainly going up all over the world, but whether intelligence itself has risen remains controversial. American Scientist. 1997; 85 :440–47. [ Google Scholar ]
• Pietschnig Jakob, Gittler Georg. A reversal of the Flynn effect for spatial perception in German-speaking countries: Evidence from a cross-temporal IRT-based meta-analysis (1977–2014) Intelligence. 2015; 53 :145–53. doi: 10.1016/j.intell.2015.10.004. [ CrossRef ] [ Google Scholar ]
• Pietschnig Jakob, Voracek Martin. One century of global IQ gains: A formal meta-analysis of the Flynn Effect (1909–2013) Perspectives on Psychological Science. 2015; 10 :282–306. doi: 10.1177/1745691615577701. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Pietschnig Jakob, Deimann Pia, Hirschmann Nicole, Kastner-Koller Ursula. The Flynn effect in Germanophone preschoolers (1996–2018): Small effects, erratic directions, and questionable interpretations. Intelligence. 2021; 86 :101544. doi: 10.1016/j.intell.2021.101544. [ CrossRef ] [ Google Scholar ]
• Pietschnig Jakob, Oberleiter Sandra, Toffalini Enrico, Giofrè David. Reliability of the g factor over time in Italian INVALSI data (2010–2022): What can achievement-g tell us about the Flynn effect? Personality and Individual Differences. 2023; 214 :112345. doi: 10.1016/j.paid.2023.112345. [ CrossRef ] [ Google Scholar ]
• R Core Team . R: A Language and Environment for Statistical Computing. R Core Team; Vienna: 2022. [(accessed on 27 December 2023)]. Available online: https://www.R-project.org/ [ Google Scholar ]
• Rindermann Heiner, Thompson James. Ability rise in NAEP and narrowing ethnic gaps? Intelligence. 2013; 41 :821–31. doi: 10.1016/j.intell.2013.06.016. [ CrossRef ] [ Google Scholar ]
• Rosseel Yves. Lavaan: An R package for structural equation modeling. Journal of Statistical Software. 2012; 48 :1–36. doi: 10.18637/jss.v048.i02. [ CrossRef ] [ Google Scholar ]
• Schaie K. Warner, Strother Charles R. A cross-sequential study of age changes in cognitive behavior. Psychological Bulletin. 1968; 70 :671–80. doi: 10.1037/h0026811. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Schneider W. Joel, McGrew Kevin S. Contemporary Intellectual Assessment: Theories, Tests, and Issues. Guilford Press; New York: 2018. The Cattell-Horn-Carroll theory of cognitive abilities; pp. 73–163. [ Google Scholar ]
• Spearman Charles. “General intelligence” objectively determined and measured. American Journal of Psychology. 1904; 15 :201–93. doi: 10.2307/1412107. [ CrossRef ] [ Google Scholar ]
• Wilhelm Oliver, Schroeders Ulrich, Schipolowski Stefan. BEFKI 8–10: Berliner Test zur Erfassung Fluider und Kristalliner Intelligenz für die 8. bis 10. Jahrgangsstufe; Manual [BEFKI 8–10: Berlin Test for the Assessment of Fluid and Crystallized Intelligence for Grades 8 to 10: Manual] Hogrefe; Boston: 2014. [ Google Scholar ]
• Woodley Michael A., Nijenhuis Jan te, Must Olev, Must Aasa. Controlling for increased guessing enhances the independence of the Flynn effect from g : The return of the Brand effect. Intelligence. 2014; 43 :27–34. doi: 10.1016/j.intell.2013.12.004. [ CrossRef ] [ Google Scholar ]
• Wu Hao, Estabrook Ryne. Identification of confirmatory factor analysis models of different levels of invariance for ordered categorical outcomes. Psychometrica. 2016; 81 :1014–1045. doi: 10.1007/s11336-016-9506-0. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]

Neurodiversity and Intelligence: Evaluating the Flynn Effect in Children with Autism Spectrum Disorder

• Original Article
• Published: 03 May 2021
• Volume 53 , pages 919–927, ( 2022 )

Cite this article

• Kenzie B. Billeiter   ORCID: orcid.org/0000-0002-4894-7127 1 ,
• John Mark Froiland   ORCID: orcid.org/0000-0002-2700-3987 2 ,
• Justin P. Allen   ORCID: orcid.org/0000-0003-0172-5970 3 &
• Daniel B. Hajovsky   ORCID: orcid.org/0000-0002-2792-2289 4  

1677 Accesses

63 Altmetric

Explore all metrics

The Flynn Effect (FE) among child and adolescent populations indicates that intelligence scores improve by about three points per decade. Using nine years of data from the National Database for Autism Research, this study examined whether general intelligence changed significantly for nine cohorts with autism spectrum disorder (ASD; N  = 671). Analyses demonstrated a downward trend such that Cohen’s d from 1998 to 2006 was − 0.27. The mean IQ is 92.74 for years 1–3, 91.54 for years 4–6, and 87.34 for years 7–9, indicating a reverse FE of 5.4 points per decade. A linear regression revealed a significant negative FE comparable to the positive effect of age on IQ among those with ASD. Implications for research, practice, and law are discussed.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

Embryonic origin of two ASD subtypes of social symptom severity: the larger the brain cortical organoid size, the more severe the social symptoms

The gift of dyslexia: what is the harm in it?

The Diagnosis of Autism: From Kanner to DSM-III to DSM-5 and Beyond

Flynn JR (2006) Beyond the Flynn effect: solution to all outstanding problems - except enhancing wisdom. http://www.thepsychometricscentre.co.uk/publications/BeyondTheFlynnEffect.asp

Shenk D (2017) What is the Flynn effect, and how does it change our understanding of IQ? WIREs Cogn Sci 8:1–2

Article   Google Scholar  

Trahan LH, Stuebing KK, Hiscock M, Fletcher JM (2014) The Flynn effect: a meta-analysis. Psychol Bull 140:1332–1360

Article   PubMed   PubMed Central   Google Scholar  

Kaufman AS, Dillon T, Kirsch JW (2013) A beautiful theory, killed by a nasty, ugly little fact. Psyccritiques 58:19

Google Scholar  

Wechsler D (2008) Wechsler adult intelligence scale–fourth edition. Pearson, United States

Kaufman AS (2010) “In what way are apples and oranges alike?” A critique of Flynn’s interpretation of the Flynn effect. J Psychoeduc Assess 28:382–398

McGrew KS (2010) The Flynn effect and its critics: rusty Linchpins and “Lookin” for g and G f in some of the wrong places’. J Psychoeduc Assess 28:448–468

Hajovsky D, Reynolds MR, Floyd RG, Turek JJ, Keith TZ (2014) A multigroup investigation of latent cognitive abilities and reading achievement relations. Sch Psychol Rev 43:385–406

Villeneuve EF, Hajovsky DB, Mason BA, Lewno BM (2019) Cognitive ability and math computation developmental relations with math problem solving: an integrated, multigroup approach. Sch Psychol 34:96

Hajovsky DB, Villeneuve EF, Reynolds MR, Niileksela CR, Mason BA, Shudak NJ (2018) Cognitive ability influences on written expression: evidence for developmental and sex-based differences in school-age children. J Sch Psychol 67:104–118

Article   PubMed   Google Scholar  

Froiland JM, Davison ML (2020) Social perception: relationships with general intelligence, working memory, processing speed, visual-spatial ability, and verbal comprehension. Ed Psychol 40:750–766

Danner D, Hagemann D, Schankin A, Hager M, Funke J (2011) Beyond IQ: a latent state-trait analysis of general intelligence, dynamic decision making, and implicit learning. Intelligence 39:323–334

Pietschnig J, Voracek M (2015) One century of global IQ gains: a formal meta-analysis of the Flynn effect (1909–2013). Perspect Psychol Sci 10:282–306

Carl N (2016) IQ and socioeconomic development across regions of the UK. J Biosocl Sci 48:406–417

Marks DF (2010) IQ variations across time, race, and nationality: an artifact of differences in literacy skills. Psychol Rep 106:643–664

Colom R, Lluis-Font JM, Andrés-Pueyo A (2005) The generational intelligence gains are caused by decreasing variance in the lower half of the distribution: supporting evidence for the nutrition hypothesis. Intelligence 33:3–91

Lynn R (1998) In support of the nutrition theory. In: Nesser U (ed) The rising curve: long-term gains in IQ and related measures. American Psychological Association, Washington, DC, pp 207–215

Chapter   Google Scholar  

Mingroni MA (2007) Resolving the IQ paradox: heterosis as a cause of the Flynn effect and other trends. Psychol Bull 114:806–829

Woodley M, Meisenberg G (2013) In the Netherlands the anti-Flynn effect is a Jensen effect. Pers Individ Differ 54:871–876

Dutton E, van der Linden D, Lynn R (2016) The negative Flynn effect: a systematic literature review. Intelligence 59:163–169

Sundet JM, Barlaug DG, Torjussen TM (2004) The end of the Flynn effect? A study of secular trends in mean intelligence test scores of Norwegian conscripts during half a century. Intelligence 32:349–362

Teasdale T, Owen DR (2008) Secular declines in cognitive test scores: a reversal of the Flynn effect. Intelligence 36:21–126

Shayer M, Ginsburg D (2007) Thirty years on—a large anti-Flynn effect? The Piagetian test volume and heaviness norms 1975–2003. J Educ Psychol 77:25–41

Shayer M, Ginsburg D (2009) Thirty years on—a large anti-Flynn effect? 11: 13-year olds. piagetian tests of formal operations norms 1976–2006/7. Br J Educ Psychol 79:409–418

Dutton E, Lynn R (2013) A negative Flynn effect in Finland, 1997–2009. Intelligence 41:817–820

Dutton E, Lynn R (2015) A negative Flynn effect in France, 1999–2008/9. Intelligence 51:67–70

Korgesaar M (2013) Flynni Efekti Esinemine Eesti Abiturientide Seas Raveni Testi Pohjal. (Seminaritöö) Tartu Ulikool

Platt JM, Keyes KM, McLaughlin KA, Kaufman AS (2019) The Flynn effect for fluid IQ may not generalize to all ages or ability levels: a population-based study of 10,000 US adolescents. Intelligence 77:101385

Bain GH, Lemmon H, Teunisse S, Starr J, Fox HC, Deary IJ, Whalley LJ (2003) Quality of life in healthy old age: relationships with childhood IQ, minor psychological symptoms and optimism. Soc Psychiatry Psychiatr Epidemiol 38:632–636

Rajput S, Hassiotis A, Richards M, Hatch SL, Stewart R (2011) Associations between IQ and common mental disorders: the 2000 British National Survey of Psychiatric Morbidity. Eur Psychiatry 26:390–395

Flynn JR (2006) Tethering the elephant: capital cases, IQ, and the Flynn effect. Psychol Pub Pol’y & L 12:170–189

Kanaya T, Ceci S (2012) The impact of the Flynn effect on LD diagnoses in special education. J Learn Disabil 45:319–326

Wechsler D (1974) Manual for the Wechsler intelligence scale for children-revised. The Psychological Corporation, San Antonio, Texas

Wechsler D (1991) Wechsler Intelligence Scale for children: third edition manual. The Psychological Corporation, San Antonio, Texas

Matson JL, Williams LW (2013) Differential diagnosis and comorbidity: distinguishing autism from other mental health issues. Neuropsychiatry 3:233–243

O’Riordan MA, Plaisted KC, Driver J, Baron-Cohen S (2001) Superior visual search in autism. Journal Exp Psychol Hum Percept Perform 27:719

Pellicano E, Burr D (2012) When the world becomes ‘too real’: a Bayesian explanation of autistic perception. Trends Cogn Styles 16:504–510

Remington A, Fairnie J (2017) A sound advantage: Increased auditory capacity in autism. Cognition 166:459–465

Shah A, Frith U (1993) Why do autistic individuals show superior performance on the block design task? J Child Psychol Psychiatry 34:1351–1364

Tillmann J, Swettenham J (2017) Visual perceptual load reduces auditory detection in typically developing individuals but not in individuals with autism spectrum disorders. Neuropsychology 31:181–190

Baron-Cohen S, Wheelwright S, Skinner R, Martin J, Clubley E (2001) The Autism-Spectrum Quotient (AQ): evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. J Autism Dev Disord 31:5–17

Happé F (1999) Autism: cognitive deficit or cognitive style? https://www.ncbi.nlm.nih.gov/pubmed/10354574

Coolican J, Bryson SE, Zwaigenbaum L (2008) Brief report: Data on the Stanford-Binet intelligence scales in children with autism spectrum disorder. J Autism Dev Disord 38:190–197

Lennen DT, Lamb GD, Dunagan BJ, Hall TA (2010) Verbal prowess equals higher IQ: implications for evaluating autism. Res Autism Spectr Disord 4:95–101

Crespi BJ (2016) Autism as a disorder of high intelligence. Front Neurosci 10:300

Oliveras-Rentas RE, Kenworthy L, Roberson RB, Martin A, Wallace GL (2012) WISC-IV profile in high-functioning autism spectrum disorders: impaired processing speed is associated with increased autism communication symptoms and decreased adaptive communication abilities. J Autism Dev Disord 42:655–664

Koegel LK, Koegel RL, Smith A (1997) Variables related to differences in standardized test outcomes for children with autism. J Autism Dev Disord 27:233–244

Freeman BJ, Del’Homme M, Guthrie D, Zhang F (1999) Vineland Adaptive Behavior Scale scores as a function of age and initial IQ in 210 autistic children. J Autism Dev Disord 29:379–384

American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders, 5th edn. APA, Washington, DC

Book   Google Scholar  

Wechsler D (1992) Wechsler preschool and primary scale of intelligence, 3rd edn. The Psychological Corporation, San Antonio, TX

Wechsler D (1991) The Wechsler intelligence scale for children—third edition. The Psychological Corporation, San Antonio, TX

Wechsler D (1981) Wechsler adult intelligence scale, Revised. The Psychological Corporation, San Antonio, TX

Bishop SL, Farmer C, Thurm A (2015) Measurement of nonverbal IQ in autism spectrum disorder: scores in young adulthood compared to early childhood. J Autism Devl Disord 45:966–974

Solomon M, Iosif AM, Reinhardt VP, Libero LE, Nordahl CW, Ozonoff S et al (2018) What will my child’s future hold? Phenotypes of intellectual development in 2–8-year-olds with autism spectrum disorder. Autism Res 11:121–132

Wechsler D (1999) Wechsler abbreviated scale of intelligence. Pearson, London

Rogers SJ, Dawson G (2020) Early Start Denver Model for young children with autism: promoting language, learning, and engagement. Guilford Publications, New York

Eaves LC, Ho HH (2004) The very early identification of autism: outcome to age 41/2–5. J Autism Dev Disord 34:367–378

Yang P, Jong YJ, Hsu HY, Chen CS (2003) Preschool children with autism spectrum disorders in Taiwan: follow-up of cognitive assessment to early school age. Brain Dev 25:549–554

Mayes SD, Calhoun SL (2003) Ability profiles in children with autism: influence of age and IQ. Autism 7:65–80

Lynn R, Hampson S (1986) The rise of national intelligence: evidence from Britain, Japan and the USA. Pers Individ Differ 7:23–32

Spitz HH (1989) Variations in Wechsler interscale IQ disparities at different levels of IQ. Intelligence 13:157–167

Teasdale TW, Owen DR (1989) Continuing secular increases in intelligence and stable prevalence of high intelligence levels. Intelligence 13:255–262

Zhou X, Zhu J, Weiss LG (2010) Pearson Peeking inside the “blackbox” of the Flynn effect: evidence from three Wechsler instruments. J Psychoeduc Assess 28:399–411

Sattler JM (1992) Assessment of children: Revised and updated

Kezer F, Arik RS (2012) An examination and comparison of the revisions of the Wechsler Intelligence Scale for children. Procedia Soc Behav Sci 46:2104–2110

Becker LA (1999) Effect size calculators. https://www.uccs.edu/lbecker/

Landa RJ (2018) Efficacy of early interventions for infants and young children with, and at risk for, autism spectrum disorders. Int Rev Psychiatry 30:25–39

Crone EA, Donohue SE, Honomichl R, Wendelken C, Bunge SA (2006) Brain regions mediating flexible rule use during development. J Neurosci 26:11239–11247

Crone EA, Wendelken C, Van Leijenhorst L, Honomichl RD, Christoff K, Bunge SA (2009) Neurocognitive development of relational reasoning. Dev Sci 12:55–66

Ferrer E, O’Hare ED, Bunge SA (2009) Fluid reasoning and the developing brain. Front Neurosci 3:3

Lövdén M, Fratiglioni L, Glymour MM, Lindenberger U, Tucker-Drob EM (2020) Education and cognitive functioning across the life span. Psychol Sci Publ Int 21:6–41

Shaw SR, Jankowska AM (2018) Pediatric intellectual disabilities at school: translating research into practice. Springer, New York

Szatmari P (2018) Risk and resilience in autism spectrum disorder: a missed translational opportunity? Dev Med Child Neurol 60:225–229

Orinstein AJ, Suh J, Porter K et al (2015) Social function and communication in optimal outcome children and adolescents with an autism history on structured test measures. J Autism Dev Disord 45:2443–2463

Sigman M, McGovern CW (2005) Improvement in cognitive and language skills from preschool to adolescence in autism. J Autism Dev Disord 35:15–23

Froiland JM, Davison ML (2020) Emotional intelligence, listening comprehension, and reading comprehension among diverse adolescents. J Child Fam Stud 29:1385–1390

Mayer JD, Caruso DR, Salovey P (2016) The ability model of emotional intelligence: principles and updates. Emot Rev 8:290–300

Reichow B (2012) Overview of meta-analyses on early intensive behavioral intervention for young children with autism spectrum disorders. J Autism Dev Disord 42:512–520

Smith T, Iadarola S (2015) Evidence base update for autism spectrum disorder. J Clin Child Adolesc Psychol 44:897–922

Benson N, Beaujean AA, Taub GE (2015) Using score equating and measurement invariance to examine the Flynn effect in the Wechsler adult intelligence scale. Multivar Beh Res 50:398–415

Download references

Acknowledgment

Data and/or research tools used in the preparation of this manuscript were obtained from the National Institute of Mental Health (NIMH) Data Archive (NDA). NDA is a collaborative informatics system created by the National Institutes of Health to provide a national resource to support and accelerate research in mental health. Dataset identifier(s): Package number 113988. This manuscript reflects the views of the authors and may not reflect the opinions or views of the NIH or of the Submitters submitting original data to NDA.

Author information

Authors and affiliations.

Department of Educational Psychology, Baylor University, Waco, TX, USA

Kenzie B. Billeiter

Department of Educational Studies, Purdue University, West Lafayette, IN, USA

John Mark Froiland

Department of Psychology, Sam Houston State University, Huntsville, TX, USA

Justin P. Allen

Division of Counseling and Psychology in Education, University of South Dakota, Vermillion, SD, USA

Daniel B. Hajovsky

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Kenzie B. Billeiter .

Ethics declarations

Conflict of interest.

We have no conflict of interest to disclose.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Billeiter, K.B., Froiland, J.M., Allen, J.P. et al. Neurodiversity and Intelligence: Evaluating the Flynn Effect in Children with Autism Spectrum Disorder. Child Psychiatry Hum Dev 53 , 919–927 (2022). https://doi.org/10.1007/s10578-021-01175-w

Download citation

Accepted : 21 April 2021

Published : 03 May 2021

Issue Date : October 2022

DOI : https://doi.org/10.1007/s10578-021-01175-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Intelligence
• Neurodiversity
• Autism Spectrum Disorder
• Flynn effect
• Find a journal
• Publish with us
• Track your research

The Flynn Effect – Explaining Increasing IQ Scores

Eleanor Myers

Lab Manager at Duke University

Psychology Major at Princeton University

Eleanor Myers is a Princeton University psychology graduate.  At Princeton Eleanor studied language development as a research assistant in the Princeton Baby Lab. Eleanor is interested in how atypical child populations learn language, and how social cues and interactions can aid in language development. Eleanor currently works as a lab manager of the early childhood cognition lab at Duke University.

Learn about our Editorial Process

Saul Mcleod, PhD

Editor-in-Chief for Simply Psychology

BSc (Hons) Psychology, MRes, PhD, University of Manchester

Saul Mcleod, PhD., is a qualified psychology teacher with over 18 years of experience in further and higher education. He has been published in peer-reviewed journals, including the Journal of Clinical Psychology.

On This Page:

The Flynn Effect refers to the finding that the average human IQ has increased over time, first discovered by researcher James Flynn in 1984.

The term “Flynn Effect” was created by Herrnstein and Murray in 1994 to refer to James Flynn’s findings of this increase in IQ over time (Williams, 2013; Herrnstein & Murray, 2010).

In his groundbreaking paper, Flynn found evidence “that representative samples of Americans did better and better on IQ tests over 46 years, the total gain amounting to a rise in mean IQ of 13.8 points” (Flynn, 1984).

With the assumption that IQ tests accurately represent intelligence , this result indicates an increase in human intelligence over time.

As Flynn discussed in his Ted Talk in 2013, human civilization has seen a big increase in IQ over time, as every generation gets more and more questions right on IQ tests (Flynn, 2013). 

He explains that “if you score the people a century ago against modern norms, they would have an average IQ of 70.  If you score us against their norms, we would have an average IQ of 130” (Flynn, 2013).

Interestingly, there has been the biggest increase in IQ in specific areas of the test: classification and analogies (being able to use logic on abstractions) (Flynn, 2013). 

Flynn explains that this is likely representative of the change in thinking patterns of humans over time, especially when it comes to the hypothetical (Flynn, 2013).

Why Does the Flynn Effect Happen?

There are a variety of explanations for the Flynn Effect. James Flynn himself describes some of these possible explanations in his Ted Talk (Flynn, 2013).

First, he discusses education as an explanation, describing how “the tenor of education has changed.  We are educating people to take the hypothetical seriously, to use abstractions, and to link them logically” (Flynn, 2013).

This development in education likely explains the observed increase in IQ specifically on analogies and being able to apply logic to abstract ideas (Flynn, 2013).

In addition to higher quality education being available in modern times, more people have access to education now than in the past, which may also contribute to the effect.

Access to information

We are in an age of information and can research almost any topic through the Internet. It is much quicker and easier to digest information than it would have been in past decades. Thus, with more access to information, it makes sense to believe our intelligence can increase.

Exposure to complex tasks

Over time, our world has become more and more complex, with new inventions and developments occurring all around us. These new aspects of our world are often much more advanced and require a developed mind to engage with and understand them.

For example, the relatively new emergence of video and computer games has led to humans developing skills in these cognitively demanding tasks. Therefore, exposure to new, more advanced tasks may be another explanation for the Flynn Effect.

Health and nutrition

Finally, it is thought that improvements in health and nutrition may both help to explain the Flynn Effect.

Over the past century, there has been a better awareness of health, such as there being a decrease in the number of people who smoke and the discontinuation of the use of harmful lead paint. There have also been improvements in the prevention and treatment of infectious diseases and improvements in nutrition.

Having healthier people can mean that more individuals can reach their full potential and become more intelligent.

Why is the Flynn Effect Important?

The Flynn Effect is important because it highlights the development of human intelligence over time. Although it may be obvious in certain areas that human civilization has become more intelligent with time, the Flynn Effect provides a concrete representation of this increase.

Using IQ tests to look at changes in human intelligence over time provides a tangible representation of intelligence that can be compared and contrasted.

Additionally, the Flynn Effect suggests that modern education works and is effective. The Flynn Effect may therefore encourage more modern teaching approaches.

If IQ tests accurately represent human intelligence, then an increase in the average human IQ score over time would suggest an increase in intelligence.

Intelligence can be developed and increased with education, suggesting that recent education methods may be more effective than past methods.

Finally, another reason that the Flynn Effect is so important is because of the role of IQ in intellectual disabilities, especially “in high stakes decisions when an IQ cut point is used as a necessary part of the decision-making process,” such as in “the determination of intellectual disability in capital punishment cases” (Trahan et al., 2014).

Additionally, IQ plays a role in “determining eligibility for special education and American Disability Act services and Social Security Disability Insurance (SSDI) in the United States” (Trahan et al., 2014).

These are just a few examples of how important IQ can be in determining the course of some people’s lives. This emphasizes the importance of research on human IQ and how it may change over time.

Is the Flynn Effect Valid?

Supporting evidence.

The 2014 meta-analysis of the Flynn Effect, as mentioned above, perhaps provides the most support for the Flynn Effect due to the size of the study (Trahan et al., 2014).  This study found that “across 285 studies (N = 14,031) since 1952 with administrations of 2 intelligence tests with different normative bases, the meta-analytic mean was 2.31…standard score points per decade” (Trahan et al., 2014).

This meta-analysis provides support for the Flynn Effect on a wide scale, given the large number of studies that it used in its analysis (Trahan et al., 2014).

In addition to James Flynn’s original study of the Flynn Effect, he also conducted a study in 1987 where he found that “data from 14 nations reveal IQ gains ranging from 5 to 25 points in a single generation” (Flynn, 1987, as cited in Rodgers, 1998).

These findings provide additional support for the Flynn Effect in a variety of places around the world (Flynn, 1987, as cited in Rodgers, 1998; (Flynn, 2009, as cited in Trahan et al., 2014).

Conversely, there are a variety of studies and reviews that critique the Flynn Effect.

One such review by Rodgers (1998)  argues that “the acceptance of the effect has been too quick” and that “before the effect is taken seriously by the community of social science researchers, its very existence should not be questionable” (Rodgers, 1998).

Furthermore, Rodgers argues that “research addressing the legitimacy and meaning of the effect should precede research testing for and evaluating causes of the effect” (Rodgers, 1998).

Because there are opposing views on the existence of the Flynn Effect and other aspects of the effect, research should focus on finding a more solid foundation of the effect before exploring other interactions of the effect (Rodgers, 1998).

Overall, more data on IQ tests from more participants from a variety of places around the world will be beneficial in achieving this goal.

Is the Flynn Effect Reversing?

Some researchers have found evidence for a “negative Flynn Effect,” that is, that the Flynn Effect has actually begun to reverse (Dutton et al., 2016, as cited in Bratsberg & Rogeberg, 2018; Pietschnig & Voracek, 2015, as cited in Bratsberg & Rogeberg, 2018)

Numerous studies on the Flynn Effect in a variety of countries have provided support for this “negative Flynn Effect” (Dutton et al., 2016, as cited in Bratsberg & Rogeberg, 2018).

However, a 2014 Flynn Effect meta-analysis did not support the Flynn Effect reversing (Trahan et al., 2014).

There seems to be support for both arguments, which emphasizes the need for further research on the Flynn Effect to determine what is the true pattern of human IQ throughout history.

What is the Flynn effect?

The Flynn Effect refers to the substantial and consistent rise in average IQ scores observed over the past century in numerous countries, as political scientist James Flynn discovered. This increase is attributed to environmental factors like improved nutrition, education, and reduced exposure to toxins.

Why does the Flynn effect occur?

The likely cause of the Flynn effect, characterized by rising performance on IQ tests over the past century, is thought to be multifactorial and largely due to environmental changes. These factors include improved nutrition, more and better education, reduced exposure to toxins, and increased cognitive stimulation from technology.

Furthermore, societal changes that emphasize more complex cognitive skills may have contributed. It’s important to note that while IQ scores have increased, it doesn’t necessarily mean inherent intelligence has changed, rather our abilities to solve certain problems have improved.

What is an example of the Flynn Effect?

An example of the Flynn Effect is in intelligence scores. It is thought that if a person took an IQ test in the 19th Century, the average score would be significantly lower than it would be if that same person took an IQ test today.

This is because the average human IQ is believed to have increased over time, and therefore someone would naturally perform better on an IQ test nowadays than the same person would perform on an IQ test decades ago.

How does the Flynn effect relate to education?

The Flynn Effect relates to education because education is often thought of as relating to IQ levels. It is a common thought that a human that has been able to access education would have a higher IQ than a human who has not been able to access education.

Additionally, someone who has been able to receive a higher level of education than someone else would likely be expected to have a higher IQ. And, as James Flynn explained in his Ted Talk, education is a key possible explanation for the Flynn Effect (Flynn, 2013).

Are jobs more cognitively demanding than they were in the past?

As James Flynn discussed in his Ted Talk , jobs have become more cognitively demanding over time (Flynn, 2013).

This development has both contributed to the increasing complexity of the world around us, and has also been a result of this increase. More people with more cognitively demanding jobs has led to the development of many advanced aspects of our world today, including new technologies and inventions.

Furthermore, these new technologies and inventions have led to the creation of more cognitively demanding jobs, as they require more people to be able to work with and further develop these new advances in our society.

Are IQ tests reliable?

It is important to consider the reliability of IQ tests and how well, if at all, they truly represent the intelligence of human beings.

James Flynn discusses the reliability of IQ tests in his 1987 paper titled “ Massive IQ Gains in 14 Nations: What IQ tests really measure” (Flynn, 1987).

In this paper, Flynn looked at increases in IQ in 14 nations and concluded that “IQ tests do not measure intelligence but rather a correlate with a weak causal link to intelligence” (Flynn, 1987).

It is also questionable whether IQ tests can be used to measure intelligence in various cultures. IQ tests are argued to be culturally biased since they are often designed for and favor white, middle-class groups and may not apply to other groups (Ford, 2004). It is important to consider what IQ tests measure when attributing the measurement of human intelligence to these tests so that we can draw accurate conclusions about human intelligence overall.

Bratsberg, B., & Rogeberg, O. (2018). Flynn effect and its reversal are both environmentally caused. Proceedings of the National Academy of Sciences, 115 (26), 6674-6678.

Dutton, E., van der Linden, D., & Lynn, R. (2016). The negative Flynn Effect: A systematic literature review. Intelligence, 59 , 163-169.

Flynn, J. R. (1984). The mean IQ of Americans: Massive gains 1932 to 1978. Psychological Bulletin, 95 (1).

Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101 (2), 171.

Flynn, J. R. (2007).  What is intelligence?: Beyond the Flynn effect . Cambridge University Press.

Flynn, J. R. (2009). The WAIS-III and WAIS-IV: Daubert motions favor the certainly false over the approximately true. Applied Neuropsychology, 16 (2), 98-104.

Flynn, J. (2013). Why our IQ levels are higher than our grandparents” | James Flynn. Ted Talk https://www.youtube.com/watch?v=9vpqilhW9uI

Ford, D. Y. (2004). Intelligence testing and cultural diversity: Concerns, cautions, and considerations. National Research Center on the Gifted and Talented.

Herrnstein, R. J., & Murray, C. (2010). The bell curve: Intelligence and class structure in American life . Simon and Schuster.

Lynn, R. (2009). What has caused the Flynn effect? Secular increases in the Development Quotients of infants.  Intelligence ,  37 (1), 16-24.

Pietschnig, J., & Voracek, M. (2015). One century of global IQ gains: A formal meta-analysis of the Flynn effect (1909–2013). Perspectives on Psychological Science, 10 (3), 282-306.

Rodgers, J. L. (1998). A critique of the Flynn effect: Massive IQ gains, methodological artifacts, or both?. Intelligence, 26 (4), 337-356.

Teasdale, T. W., & Owen, D. R. (2008). Secular declines in cognitive test scores: A reversal of the Flynn Effect.  Intelligence ,  36 (2), 121-126.

Trahan, L. H., Stuebing, K. K., Fletcher, J. M., & Hiscock, M. (2014). The Flynn effect: a meta-analysis .  Psychological Bulletin, 140 (5), 1332.

Williams, R. L. (2013). Overview of the Flynn effect. Intelligence, 41(6) , 753-764.

Related Articles

Soft Determinism In Psychology

Branches of Psychology

Social Action Theory (Weber): Definition & Examples

Adult Attachment , Personality , Psychology , Relationships

Attachment Styles and How They Affect Adult Relationships

Personality , Psychology

Big Five Personality Traits: The 5-Factor Model of Personality

Learning Theories , Psychology

Behaviorism In Psychology

Intelligent Insights on Intelligence Theories and Tests

Tuesday, October 25, 2016

Research byte: the negative flynn effect: a systematic literature review via browzine.

No comments:

Post a Comment