Main study
1. Study identifies a drop in children’s intelligence scores between 2016 and 2021.
A study involving more than two thousand students found a decline in general, verbal, and visual intelligence scores.
A study published in 2023 in the Journal of Theoretical Educational Science analyzed the results of 2,192 first-grade students in Turkey. Researchers identified a downward trend in general, verbal, and visual intelligence scores between 2016 and 2021, with a sharper decline between 2020 and 2021. The results contradict the historical trend of IQ growth, known as the Flynn effect. However, the study does not demonstrate that the intellectual capacity of new generations has decreased globally, nor does it establish the causes of the observed decline.
Source: Intelligence Differences across Years: A Trend Analysis , Bal-Sezerel and collaborators, 2023.
Flynn effect and generational changes
2. Experts point to education, health, and nutrition as explanations for the historic increase in IQ.
The research gathered the opinions of 75 experts on changes in intelligence across generations.
A survey published in Personality and Individual Differences investigated how experts explain the historical growth in intelligence test scores. Among the main factors mentioned were improvements in education, nutrition, health, and living conditions. Participants also considered the possibility of stagnation or decline in scores in some regions. The projections represent assessments by the experts consulted, not proven predictions about the intelligence of future generations.
Source: Survey of Expert Opinion on Intelligence: The Flynn Effect and the Future of Intelligence , Rindermann and collaborators, 2016.
3. Scientific database gathers intelligence data from over 200,000 people over 65 years.
Research has compiled international results from reasoning tests conducted between 1950 and 2014.
Researchers gathered information from 202,468 participants to study historical changes in Raven’s Progressive Matrices test scores. Published in Data in Brief in 2020, the work provides data from hundreds of samples, allowing for comparisons across time periods, age groups, and national contexts. This material forms a basis for further analysis of the Flynn effect. The article primarily presents the data used in the investigation, not a single causal conclusion about the increase in IQ.
Source: IQ Score Gains over 65 Years Worldwide: Cross-temporal Meta-analysis Datasets , Wongupparaj and collaborators, 2020.
4. Children from rural Kenya showed an increase in IQ scores.
A comparison between 1984 and 1998 identified improvements, especially in nonverbal reasoning tasks.
A study published in the journal Psychological Science compared the results of children from the Embu region of Kenya in two periods separated by 14 years. Researchers found significant increases in intelligence test scores, especially in Raven’s Matrices. The study discussed possible influences of schooling, parental literacy, nutrition, and environmental complexity. The results demonstrate that the Flynn effect has also been observed outside industrialized countries, although they do not identify a single cause.
Source: IQ on the Rise: The Flynn Effect in Rural Kenyan Children , Daley and colleagues.
5. Researcher questions whether the historical increase in IQ represents a real growth in intelligence.
Scientific analysis identifies methodological limitations in early investigations into the Flynn effect.
Psychologist Joseph Rodgers examined studies that solidified the idea of increasing intelligence scores throughout the 20th century. His analysis acknowledged that the phenomenon cannot be explained solely by methodological errors, but questioned hasty interpretations. According to the researcher, understanding how scores change requires investigating changes in the distribution of results, the characteristics of the tests, and the populations evaluated. The work advocates for greater precision before attributing causes to the phenomenon.
Source: A Critique of the Flynn Effect: Massive IQ Gains, Methodological Artifacts, or Both? , Rodgers, 1999.
6. Research with 1.7 million results identifies cognitive gains even among high-achieving students.
The study investigated whether the Flynn effect reaches the top 5% of intellectual performers.
An analysis published in the journal Intelligence in 2011 examined approximately 1.7 million academic test results administered between 1981 and 2010. Researchers identified historical gains even among students in the upper performance range, especially in mathematics tests. The trend appeared in participants of both sexes. The findings indicate that the increase in scores did not occur only among average-achieving students, although the tests used do not represent a comprehensive measure of intelligence.
Source: The Flynn Effect Puzzle: A 30-year Examination from the Right Tail of the Ability Distribution Provides Some Missing Pieces , Wai and Putallaz, 2011.
7. Children showed gains of up to 15 points on intelligence tests after age 25.
The investigation identified distinct improvements according to socioeconomic conditions and assessed skills.
Researchers at Leiden University compared children assessed using the same instrument at intervals separated by approximately 25 years. The results indicated increases of at least 15 points in the highest socioeconomic group and eight points in the middle group. The greatest changes occurred in spatial and analytical skills, followed by memory and verbal abilities. The study suggests that different cognitive components may respond unevenly to environmental and educational changes.
Source: Children Becoming More Intelligent: Can the Flynn Effect Be Generalized to Other Child Intelligence Tests? , Resing and Tunteler, 2007.
8. The expansion of education may have contributed to the historical increase in intelligence.
Researchers have linked schooling, brain development, and reasoning skills.
A study published in the journal Intelligence in 2015 investigated the possibility that the expansion of formal education contributed to the Flynn effect. The research combined neuroimaging evidence, comparisons of schooling in agrarian communities, and historical analysis of mathematics curricula. The results indicated relationships between school experiences and cognitive processes used in abstract reasoning. The researchers argue that education is a possible partial explanation for the historical increase in scores, without excluding other environmental factors.
Source: The Cognitive Impact of the Education Revolution: A Possible Cause of the Flynn Effect on Population IQ , Baker and colleagues, 2015.
9. Study identifies growth in children’s intelligence scores in China.
A comparison of results between 1984 and 2006 found an average increase of 4.53 IQ points.
A study published in the journal Intelligence in 2012 compared test results administered to five- and six-year-old Chinese children in 1984 and 2006. Researchers identified an average increase of 4.53 points on the overall scale, with improvements in verbal and performance components. This growth corresponded to approximately 2.06 points per decade. The results reinforce the existence of generational changes in cognitive tests, but do not determine which factors produced the gains.
Source: An Increase of Intelligence Measured by the WPPSI in China, 1984–2006 , 2012.
10. Study in Estonia identifies increase in IQ, but warns of changes in the meaning of the tests.
A comparison of three generations showed that the scores are not always directly equivalent.
A study of 2,173 students in Estonia compared assessments conducted in the 1930s, 1990s, and 2000s. The authors identified historical increases in scores, including between the two most recent cohorts. However, psychometric analysis showed that the structure and meaning of some assessed components did not remain completely stable. The study warns that generational differences may reflect both changes in performance and alterations in the interpretation of the instruments.
Source: Comparability of IQ Scores over Time , Must et al., 2009.
11. The increase in IQ in the Netherlands may not represent an equivalent growth in general intelligence.
The research examined the relationship between historical gains and the overall intelligence factor.
Researchers analyzed results from two cognitive batteries used in the Netherlands to verify whether the Flynn effect corresponded to changes in the general intelligence factor, known as the g factor. The results found small to moderate negative relationships between historical gains and g factor loadings. The research suggests that increases in test scores do not necessarily represent proportional improvements in all cognitive abilities.
Source: The Secular Rise in IQs in the Netherlands: Is the Flynn Effect on g? , te Nijenhuis and van der Flier, 2007.
12. Researchers question the mathematical model used to explain the Flynn effect.
Analysis reveals inconsistencies between theoretical predictions and the historical distribution of IQ scores.
In 2002, David Rowe and Joseph Rodgers published a critique of the model proposed by Dickens and Flynn to explain the historical increase in intelligence scores. According to the authors, certain configurations of the model would imply substantial changes in the variability of the results, something that did not appear in the same proportion in the evidence examined. The investigation does not deny the Flynn effect, but questions whether the proposed mathematical mechanism can adequately reproduce the historical data.
Source: Expanding Variance and the Case of Historical Changes in IQ Means: A Critique of Dickens and Flynn (2001) , Rowe and Rodgers, 2002.
13. Research identifies gains in cognitive performance across generations of Turkish children.
A comparison conducted after three decades found progress, especially in previously disadvantaged groups.
A study published in 2011 compared the results of a human figure drawing test administered to Turkish students in 1977 and 2010. Scores increased over the three decades, with relatively greater gains among children from initially less privileged groups. However, differences between urban and rural areas did not disappear. The results show how performance on cognitive instruments can keep pace with social and educational changes.
Source: Generational Gains on the Draw-a-Person IQ Scores: A Three-decade Comparison from Turkey , Kagitcibasi and Biricik, 2011.
14. Meta-analysis confirms a historic increase of approximately 2.3 IQ points per decade.
A quantitative review compiled 285 studies on generational changes in intelligence testing.
A meta-analysis published in the Psychological Bulletin in 2014 compiled 285 studies and information from 14,031 participants. Researchers estimated an average gain of 2.31 IQ points per decade in the comparisons analyzed. In a subset of modern Wechsler and Stanford-Binet tests, the increase approached three points per decade. The results support the historical presence of the Flynn effect, but do not allow us to assume that the same trend continues in all countries and periods.
Source: The Flynn Effect: A Meta-Analysis , Trahan et al., 2014.
15. Teachers from Asian countries noticed cognitive changes that went unnoticed in Australia.
The study compared educators’ perceptions of students’ intellectual development.
Researchers interviewed experienced teachers in Australia, South Korea, and Singapore to investigate possible changes in students’ intellectual capacity. Most Australian teachers did not perceive a significant increase, while many professionals from the Asian countries reported students who were cognitively more developed. The results suggest differences in the perception of educational transformations. Because they rely on professional reports, they do not constitute a direct measurement of intelligence growth.
Source: Evidence That Rising Population Intelligence Is Impacting in Formal Education , Cocodia and collaborators, 2003.
Intelligence and school performance
16. Study links national IQ indicators to school performance in 56 countries.
Research has found a correlation between aggregate measures of intelligence and international scores in math, reading, and science.
A study published in 2009 compared national IQ estimates with the results of the 2006 PISA assessment in 56 countries. The authors reported a correlation of 0.84 between the indicators, interpreting the result as evidence of an association between cognitive and educational performance. However, national IQ estimates constructed from heterogeneous samples have limitations in terms of representativeness and comparability. The correlation does not demonstrate that intelligence is the cause of educational differences between countries.
Source: National IQs Predict Educational Attainment in Math, Reading and Science across 56 Nations , Lynn and Mikk, 2009.
17. Intelligence shows a consistent relationship with performance in mathematics and English.
A longitudinal study followed 637 Nigerian students over four years.
A study published in 2016 evaluated the relationship between psychometric intelligence and academic performance of students in Nigeria. The research followed 637 adolescents between approximately 14 and 17 years old and identified statistically significant associations between IQ, mathematics, and English language. It also found relative stability of cognitive results over the period. The conclusions indicate predictive value of intelligence assessments, but do not exclude the influence of school, family, and social conditions.
Source: Intelligence and Academic Achievement: A Longitudinal Survey , Kpolovie, 2016.
18. Research compares four intelligence tests and their ability to predict school grades.
Results show differences between instruments used to assess children.
Researchers at the University of Basel followed children initially assessed by four intelligence tests. Of the initial 103 participants, 54 had school data available three years later. All instruments showed the ability to predict overall averages, but differed in their specific prediction of mathematics and language. The study indicates that the choice of instrument can influence the usefulness of the assessment. The authors themselves highlighted the small size of the longitudinal sample.
Source: The Predictive Validity of Four Intelligence Tests for School Grades: A Small Sample Longitudinal Study , Gygi and collaborators, 2017.
19. Study with over 70,000 children identifies a strong relationship between intelligence and school performance.
A British study followed students from ages 11 to 16 and analyzed 25 subjects.
A study published in the journal Intelligence followed more than 70,000 children in England. Researchers found a correlation of 0.81 between latent measures of general intelligence at age 11 and academic performance at age 16. The association varied by subject, being particularly strong in mathematics. The results demonstrate that cognitive ability is a relevant indicator of school performance, although it is not sufficient to individually explain all educational outcomes.
Source: Intelligence and Educational Achievement , Deary et al., 2007.
20. Study links international academic performance to indicators of cognitive development.
The authors analyzed national measures of learning and intelligence as possible components of human capital.
Gerhard Meisenberg and Richard Lynn presented an analysis relating national IQ estimates to international results in mathematics, science, and reading. The work compiled information from 168 countries and territories and indirect estimates for another 28. The authors proposed using these measures in the analysis of socioeconomic development. However, extrapolations between countries and sampling differences limit the interpretation. The data do not demonstrate innate intellectual differences between populations.
Source: Intelligence: A Measure of Human Capital in Nations , Meisenberg and Lynn.
21. Research finds a strong correlation between intelligence indicators and school outcomes in 67 countries.
The international comparison used assessments in mathematics and science.
A study published in the Journal of Biosocial Science in 2007 found a correlation of 0.92 between national estimates of IQ and academic performance of students from 67 countries. The authors interpreted the association as confirmation of the validity of the indicators used. However, comparisons of this nature depend on the quality of the samples, the equivalence of the tests, and the control of educational differences. The results do not allow for establishing a causal relationship between population intelligence and academic achievement.
Source: National IQs Predict Differences in Scholastic Achievement in 67 Countries , Lynn et al., 2007.
22. IQ indicators and results in mathematics and science show an association in an international study.
The research used data from school assessments conducted in different countries.
A study published in 2007 examined the relationship between national IQ estimates and math and science scores from the 2003 TIMSS. The analysis included sets from 25 and 46 countries, depending on age range, and found high correlations. The authors also observed a positive association between cognitive indicators and per capita income. The conclusions, however, depend on questionable population estimates and do not allow us to determine whether economic factors influence cognitive performance or vice versa.
Source: National Differences in Intelligence and Educational Attainment , Lynn and Mikk, 2007.
23. Research with Indian students finds performance differences between school models.
The research assessed intelligence and academic outcomes in public and private schools, as well as social organizations.
A study published in 2020 analyzed 150 elementary school students in India using Raven’s Colored Progressive Matrices and academic achievement tests. The results indicated differences between students from public, private, and socially affiliated institutions in subjects such as mathematics, science, and languages. The work suggests the need to consider educational contexts when interpreting performance. The sample does not allow for generalizations to the entire Indian school system.
Source: Psychometric Intelligence and Academic Achievement: A Comparative Analysis of Elementary Schools , Malhotra, 2020.
24. Intelligence can predict future school outcomes, longitudinal study indicates.
The study followed 289 students to examine the temporal relationship between IQ and learning.
Researchers analyzed the results of intelligence tests and academic performance of 289 students assessed for eligibility in special education programs. The assessments took place at two points in time, separated by approximately 2.8 years. The statistical model favored the hypothesis that initial IQ scores predict later academic performance. However, the observational structure and sample characteristics do not allow for establishing a definitive causal relationship.
Source: Psychometric Intelligence and Achievement: A Cross-lagged Panel Analysis , Watkins and collaborators, 2007.
25. A 1971 study warns of the limitations of IQ tests applied to young children.
The research examined stability and changes in verbal and nonverbal scores throughout schooling.
A scientific report from the University of Colorado analyzed intelligence assessments conducted at different school levels. The results indicated that, even when there is average stability in the indicators, individual children can show significant changes in their scores. The researcher recommended caution in interpreting group tests in the early school years and in the differences between verbal and nonverbal results before the seventh grade. The study highlights the risk of drawing definitive conclusions based on early assessments.
Source: The Stability and Change of Language and Non-Language IQ Scores , Hopkins, 1971.
Intelligence, economics, and population differences
26. University study links education, health, and fertility to national IQ indicators.
A statistical model examined associations between socioeconomic conditions and estimates of cognitive performance.
A 2014 academic paper presented at De La Salle University examined the relationship between national IQ estimates and economic and social indicators. Among five variables studied, education, health, and fertility showed statistically significant associations in the final model. The results suggest that social characteristics accompany differences in the cognitive indicators used. Because this is an observational analysis based on aggregate estimates, it is not possible to conclude that the identified factors directly cause changes in intelligence.
Source: Examining the Cognitive Capital: A Regression Analysis on the Socio-Economic Factors that Affect the National Intelligence Quotient (IQ) Scores , Diamada, 2014.
Nature of the document: university academic work, without proof of publication in a journal in the archives.
27. Bali’s population intelligence estimate exposes difficulties in international comparisons.
The study used only 50 participants and statistical adjustment procedures.
A study published in 2012 examined the performance of 50 residents of Bali, Indonesia, on a nonverbal reasoning test. Researchers applied corrections for demographic and social factors in order to estimate results for a wider population. Although they presented numerical values, the small, unrepresentative sample size and the reliance on adjustments limit the reliability of the extrapolations. The work illustrates the scientific difficulties of establishing national or regional IQ averages from limited data.
Source: Intelligence in Bali: A Case Study on Estimating Mean IQ for a Population Using Various Corrections Based on Theory and Empirical Findings , Rindermann and te Nijenhuis, 2012.
28. Analysis questions the direct relationship between a country’s average IQ and its economic wealth.
The text points out that historical and economic differences can produce misleading statistical associations.
In a 2003 academic note, Jamal Munshi examined the use of national IQ averages as an explanation for income differences between countries. The author argued that omitted variables, such as natural resources, stage of development, and historical conditions, can compromise the models analyzed. He also highlighted that international correlations do not clarify the causal direction. The text advocates caution in using these indicators to explain economic development.
Source: A Note on the Relevance of the National Average IQ , Munshi, 2003.
Nature of the document: critical note, with no identification of publication in a journal in the archive.
29. Text questions racial and cultural interpretations of intelligence tests.
Analysis argues that social conditions and educational opportunities influence the interpretation of cognitive differences.
An essay titled “IQ, Intelligence, Ethnicity, and Gender” discusses cultural limitations in the use of intelligence tests. The author questions interpretations that attribute differences between groups exclusively to biological factors and draws attention to nutrition, schooling, inequality, and living conditions. The text argues that a single score does not fully represent a person’s ability. This is a critical argument, not an experimental investigation that has directly tested these hypotheses.
Source: IQ, Intelligence, Ethnicity and Gender .
Nature of the document: essay in Word format, without identifiable authorship or scientific publication in the file.
Reversing the Flynn effect and education
30. Decline in educational indicators raises debate about a possible reversal of the Flynn effect.
Research has linked changes in PISA results to discussions about generational cognitive performance.
A 2019 article analyzed international educational outcomes and discussed evidence of a slowdown or reversal in historical gains in intelligence tests. The researcher examined PISA data from 2012 and 2015, including a selection of nine countries, and identified reductions in different school indicators. The study considers environmental and educational hypotheses, but acknowledges that academic achievement and IQ are not equivalent measures. The analysis does not demonstrate a generalized decline in human intelligence.
Source: The Reversal of the Flynn Effect and Its Reflection in the Educational Arena , Zysberg, 2019.
High abilities and assessment tools
31. Book discusses challenges in identifying gifted children.
The publication presents assessment methods and highlights the importance of recognizing different developmental profiles.
A review published in the Turkish Journal of Giftedness and Education in 2018 examined a book dedicated to identifying gifted students. The work brings together discussions on cognitive testing, observation, performance, dynamic assessment, and early identification. It also addresses children with dual exceptionality, including high abilities associated with learning difficulties or neurodevelopmental conditions. The publication highlights that no single instrument guarantees perfect identification and that educational decisions require multiple sources of information.
Source: Üstün Yeteneklilerin Tanılanması , review by Özdemir and Bal-Sezerel, 2018.
Nature of the document: book review published in a periodical.
32. Study differentiates between speed of reasoning and speed of perception in children.
Research with 211 students showed that perceptual speed and speed of reasoning do not represent exactly the same skill.
A study published in 2020 examined the relationship between a timed verbal reasoning subtest of the Anadolu-Sak scale and measures of perceptual speed. Participants included 211 children aged six and seven. The instrument showed an internal consistency of 0.84, but its correlations with perceptual tasks were small. The results support the distinction between rapidly processing simple stimuli and performing reasoning quickly, showing that different components of cognitive speed need to be assessed separately.
Source: An Investigation of the Relationship between Speed-Based Verbal Reasoning Subtest of Anadolu-Sak Intelligence Scale and Perceptual Speed Tests , Kılıçarslan, Bal-Sezerel and Sak, 2020.
33. Intelligence scale developed in Türkiye shows favorable results in identifying giftedness.
This research evaluated the validity of the Anadolu-Sak instrument in students with different abilities.
A study published in Psychological Test and Assessment Modeling in 2019 evaluated 98 students identified as highly gifted and 232 students in a comparison group. The overall score on the Anadolu-Sak scale showed the greatest ability to distinguish the groups. The authors also found significant correlations between cognitive outcomes and academic performance. The findings support the use of the instrument in the investigated context, although the validity of a scale depends on the population and the conditions of application.
Source: Validity of the Anadolu-Sak Intelligence Scale in the Identification of Gifted Students , Sak and collaborators, 2019.
34. New test seeks to identify mathematical creativity among elementary school students.
This instrument assesses the ability to formulate problems, develop conjectures, and construct proofs.
A study published in the Turkish Journal of Education presented the development of a mathematical creativity test applied to 1,129 students. The instrument uses tasks involving problem formulation, conjecture development, and proof, allowing for the evaluation of different dimensions of mathematical production. Analyses indicated favorable psychometric properties and a positive association with school grades. The proposal expands assessment beyond the mere reproduction of procedures, but requires further validation in other contexts.
Source: Mathematical Creativity Test (MCT) Development for Middle School Students , Bal-Sezerel and Sak, 2022.
35. Teachers give positive evaluation to intelligence scale developed in Türkiye.
This study investigated the acceptance and perceived usefulness of an instrument used to identify cognitive abilities.
A study published in 2020 investigated the social validity of the Anadolu-Sak scale based on the perceptions of 37 teachers involved in its application. A questionnaire with 40 items assessed characteristics such as practicality, interpretation, and usefulness. The average score was 5.26 on a six-point scale, indicating a favorable evaluation by the professionals. The result demonstrates acceptance among the studied administrators, but does not replace psychometric evidence of validity and reliability.
Source: Social Validity Study of the Anadolu-Sak Intelligence Scale (ASIS) , Tamul et al., 2020.
