The Brain That Chooses a School Isn't Ready for It Yet

The article presents neuroscientific evidence that the adolescent brain does not fully mature until around the age of 25, while the period around age 15 is, by contrast, the peak of cognitive plasticity. International studies show that the type of education (academic vs. vocational) causally affects the development of intelligence — academic secondary schools (gymnázia) in Germany demonstrably boost cognitive abilities, yet Czech multi-year gymnázia surprisingly bring no measurable added value.

Every year in the Czech Republic, roughly one hundred thousand fifteen-year-olds face one of the most consequential decisions of their lives so far: gymnázium, secondary vocational school, or apprenticeship training? Some make this decision even earlier — at the age of eleven, when applying to an eight-year gymnázium. Meanwhile, over the past two decades neuroscience has produced a disquieting finding: the part of the brain responsible for rational planning, weighing long-term consequences, and controlling impulsive reactions matures dead last — and reaches full maturity only around the age of twenty-five.

The question of whether early sorting into educational tracks is beneficial or, on the contrary, harmful for adolescents thereby shifts from the realm of pedagogical debate into the realm of neurobiology. And the answer offered by the convergence of evidence from neuroscience, psychology, the economics of education, and international comparative studies is not favorable to the defenders of the current Czech system.

This article summarizes the state of scientific knowledge in three interconnected areas: how the brain changes during adolescence and what this means for learning; whether education causally raises intelligence (and whether the type of education matters); and what consequences early sorting of pupils has for individuals and for society. It draws exclusively on published scientific studies, meta-analyses, and international datasets — not on opinions or political preferences.

A longitudinal study by Gogtay and colleagues, published in 2004 in the journal Proceedings of the National Academy of Sciences, tracked the development of grey-matter density in 13 healthy children through repeated magnetic resonance imaging over a period of up to 10 years (age range 4–21 years). It found that the cerebral cortex matures in a "back-to-front" direction — sensory areas (vision, hearing) first, association areas (planning, decision-making) last (Gogtay, N. et al., 2004, PNAS, 101(21), 8174–8179).

Later research extended this timeline further. Petanjek and colleagues (2011) demonstrated on histological material that synaptic pruning in the frontal cortex continues until the age of thirty, and Miller and colleagues (2012) showed that myelination of frontal areas in humans persists into the late twenties. The frontal cortex therefore reaches full structural maturity roughly around the age of twenty-five.

This finding is also confirmed by the ABCD study (Adolescent Brain Cognitive Development), to date the largest longitudinal project focused on brain development in adolescence. Its first wave enrolled nearly 11,900 children aged 9–10 and follows them into adulthood. Initial results show that the most intensive structural changes take place in the parietal lobe and the inner part of the frontal cortex — areas crucial for integrating information and decision-making (Casey, B. J. et al., 2018, Developmental Cognitive Neuroscience, 32, 1–144).

From the work of Giedd and colleagues (1999, Nature Neuroscience) we know that grey-matter volume in the frontal lobe peaks around the age of twelve (somewhat earlier in girls) and then declines, while white matter — the axonal connections between areas — grows continuously into the late twenties. The temporal lobe reaches its grey-matter volume peak even later, around the age of sixteen. The adolescent brain is therefore not an unfinished version of the adult brain — it is a qualitatively different organ, undergoing a massive remodeling.

One of the key processes of adolescence is synaptic pruning — the gradual elimination of excess synaptic connections. Petanjek and colleagues (2011, PNAS) demonstrated, on histological material from the human frontal cortex, that the density of dendritic spines in the frontal cortex declines continuously from childhood until the age of thirty, with their density in childhood being two to three times higher than in adults. This process increases the efficiency of neural circuits — but reduces their flexibility.

Averbeck (2022, PNAS, 119(22), e2121331119) examined this double-edged nature using a computational model: in a simulation of artificial neural networks trained on working-memory tasks, pruning of connections led to greater resilience to distracting stimuli, but at the same time to a reduced ability to learn entirely new tasks. The brain becomes faster and more precise, but less adaptable. Excessive pruning is associated with schizophrenia, insufficient pruning with autism spectrum disorders (Feinberg, 1982; Tang et al., 2014, Neuron).

The key principle of pruning — "what is not used, perishes" (in English, use it or lose it) — has a direct impact on education: synapses that are actively used are preserved and strengthened; those that are not stimulated perish. The type of cognitive activity during the period of most intensive pruning (roughly ages 12–20) thus literally shapes the architecture of the adult brain.

Myelination — the wrapping of nerve fibers in a myelin sheath — is a process that substantially increases the speed of nerve-signal transmission and shortens a neuron's recovery time between discharges. Miller and colleagues (2012, PNAS), in a comparative study of the human and chimpanzee brain, demonstrated that humans have a markedly delayed course of frontal-cortex myelination compared with other primates — the process continues into the late twenties or early thirties. This extended "window of incompleteness" is probably evolutionarily advantageous: it allows for a longer period of learning and the transmission of cultural experience.

Increasing myelination, however, simultaneously reduces malleability (plasticity). Fully myelinated circuits are fast and efficient, but difficult to remodel. The adolescent brain thus exists in a dynamic tension between rising efficiency and declining flexibility — and the educational environment determines which way this balance shifts.

Steinberg's dual-systems model (2008, Developmental Review, 28(1), 78–106) describes a key mismatch in the adolescent brain: the limbic system (the center of emotion and reward) matures earlier than the control system of the frontal cortex. Galván and colleagues (2006, Journal of Neuroscience), using functional magnetic resonance imaging, demonstrated that adolescents exhibit exaggerated activation of the nucleus accumbens (the reward center) compared with both children and adults — they are therefore biologically more prone to impulsive decisions and to overvaluing immediate gratification.

Reynolds and colleagues (2018, Cerebral Cortex, 29(3), 1116–1127) further found that mesocortical dopamine fibers grow from the striatum into the frontal cortex precisely during adolescence — the system that links motivation with rational planning is thus literally under construction during the very period when adolescents are expected to make key decisions about their future.

The popular notion that there is a sharp boundary after which certain abilities can no longer be learned has been substantially challenged by recent neuroscience. Hartshorne and colleagues (2018, Cognition), in a study involving 669,498 respondents, examined the critical period for acquiring second-language grammar. They found that the ability to reach native-level grammar declines after approximately age 17.4 — but it does not end abruptly; rather, it fades gradually. The difference between a fourteen-year-old and a twenty-year-old beginner exists, but it is not insurmountable.

For executive functions — working memory, cognitive flexibility, the ability to suppress impulsive reactions — Tervo-Clemmens and colleagues (2023), in a sample of 10,766 participants, demonstrated rapid development between ages 10 and 15 followed by stabilization around ages 18–20. Fluid intelligence peaks between ages 20 and 27; working memory improves up to approximately the age of twenty-five.

Sydnor and colleagues (2023), using imaging methods, showed that indicators of synaptic malleability in association areas of the cortex (including the frontal cortex) peak around the age of fifteen — precisely the period when the Czech education system sorts pupils into educational tracks. Fuhrmann, Knoll, and Blakemore (2015, Trends in Cognitive Sciences) and Dahl and colleagues (2018, Nature) described adolescence as a "second window of opportunity" — a sensitive period in which environmental demands are extraordinarily effective at shaping cognitive abilities.

From a neuroscientific perspective, then, the period around the age of fifteen is not the end of cognitive flexibility but its peak. The decision to narrow the focus of education comes, paradoxically, at the very moment when the brain is maximally prepared for broad learning.

The question of whether education truly raises intelligence, or whether smarter people simply study longer, has been resolved thanks to natural experiments that create external variation in the length of education independent of the pupil's abilities.

Ceci's groundbreaking work (1991, Developmental Psychology) systematically assembled six independent lines of evidence demonstrating a causal relationship: studies of interrupted schooling, the summer-vacation effect (a decline in cognitive scores over the summer), studies of late school entry, studies of early school leaving, regression discontinuity around age cutoffs, and cross-cultural comparisons of populations without formal education. He estimated the effect at 2–4 IQ points per year of schooling.

The cleanest causal evidence was provided by Brinch and Galloway (2012, PNAS). They exploited the Norwegian compulsory-education reform of 1955–1972, which gradually extended compulsory schooling from 7 to 9 years. The reform's staggered rollout across municipalities created external variation enabling an instrumental-variable estimate. Using population data from the military conscription of Norwegian men born 1950–1958, they found that each additional year of schooling in adolescence causally raised IQ by approximately 3.7 points. IQ was measured by a standardized test at age 19 (Brinch, C. N. & Galloway, T. A., 2012, PNAS, 109(2), 425–430).

The most comprehensive synthesis to date was carried out by Ritchie and Tucker-Drob (2018, Psychological Science) in a meta-analysis encompassing 142 effect sizes from 42 datasets and over 600,000 participants. They analyzed three types of quasi-experimental designs and found a consistent effect on the order of 1–5 IQ points per year of education. The effects were present for both fluid and crystallized intelligence. According to the conclusions of this analysis, education appears to be one of the most consistent and durable methods for raising cognitive abilities.

A Swedish study by Cliffordson and Gustafsson (2008, Intelligence), in a sample of 48,269 men, provided an important addition: technical and science educational tracks showed stronger effects on the corresponding cognitive tests, suggesting that the content of the curriculum shapes specific cognitive abilities — that is, that what pupils learn matters.

The key question goes beyond the mere number of years of education: does the type of education — specifically the academic versus vocational track — have a causal effect on cognitive development? The German specialist literature provides a compelling answer.

Becker, Lüdtke, Trautwein, Köller, and Baumert (2012, Journal of Educational Psychology), in a groundbreaking study, tracked 1,038 students from 49 schools from grade 7 to grade 10 within the longitudinal BIJU project. Using propensity score matching on numerous variables — prior intelligence, school performance, grades, socioeconomic background — they compared the development of intelligence (measured by the nonverbal reasoning test of the KFT) across educational tracks. Even after controlling for selection effects, they demonstrated a positive causal effect of attending a Gymnasium on the growth of intelligence. The lower tracks (Realschule, Hauptschule) did not differ from each other — only the Gymnasium showed significantly higher gains.

The authors described a "scissor effect" (in German, Schereneffekt): students in the academic track diverge more and more in their cognitive development from peers in lower tracks, even after controlling for entry abilities. The Gymnasium therefore not only admits more capable students — it literally makes them even more capable.

This finding has been confirmed by further studies. Guill, Lüdtke, and Köller (2017, Learning and Instruction), in a sample of 8,628 students from the Hamburg KESS project, confirmed a consistent positive effect of the academic track on the development of fluid intelligence over a four-year period. Traini, Kleinert, and Bittmann (2021, Research in Social Stratification and Mobility), using the nationally representative NEPS-SC3 data, demonstrated that even when the sample is restricted to students with comparable entry conditions, learning progress at the Gymnasium is higher in both reading and mathematics. The most recent confirmation, from Herrmann and Bach (2025, British Journal of Educational Psychology), using propensity-weighting methods, confirmed that non-academic educational tracks apparently impede the full development of students' abilities.

Direct imaging studies comparing the brain development of academic and vocational students do not yet exist — and this is the most significant gap in the research. However, indirect evidence from related fields offers a compelling mechanism.

First, Ritchie, Bates, and Deary (2015, Developmental Psychology), in the Lothian Birth Cohort of 1936 (n = 1,091), showed that the effect of education on cognitive tests at age 70 was not mediated by general intelligence (the g factor) but consisted of direct effects on specific skills. Broader education covering more subjects therefore produces improvements in more specific skills, thereby creating a richer cognitive repertoire.

Second, different types of learning produce structural changes in different brain regions. Zatorre, Fields, and Johansen-Berg (2012, Nature Neuroscience) documented that learning changes both grey and white matter — from the formation of new synapses and dendritic branching to myelination and changes in nerve-fiber diameter. Broad education should therefore elicit distributed structural changes across many brain networks.

Third, the concept of cognitive reserve (Stern, 2009, Neuropsychologia) identifies education as a major precondition for the brain's resilience to neurodegeneration. As many as 25% of seniors who met the full pathological criteria for Alzheimer's disease at autopsy showed no clinical impairment during life — precisely because of cognitive reserve built, among other things, through education.

Diamond and Lee (2011), in a review of interventions developing executive functions in Science, found that the most successful programs focus on the whole child — academic, social, emotional, and physical development. Narrowly focused training of individual cognitive functions is less effective than multidimensional engagement. Analogously: the broad gymnázium curriculum, which requires switching between languages, sciences, humanities, and mathematics, probably exercises executive functions (especially cognitive flexibility) better than a narrowly vocational curriculum.

Bialystok, Craik, and Luk (2012, Trends in Cognitive Sciences) demonstrated that lifelong bilingualism reorganizes brain networks and contributes to cognitive reserve — bilingual patients showed a 4–5-year delay in the onset of dementia symptoms compared with monolinguals. If managing two languages builds cognitive reserve, managing multiple academic subjects could act similarly.

Here the international evidence runs up against a peculiarity of the Czech context. Unlike German Gymnasiums, the Czech multi-year gymnáziums do not show convincing added value.

The most rigorous Czech evidence comes from the longitudinal CLoSE study (Czech Longitudinal Study in Education), which tracked approximately 6,000 students from grade 4, through the transition to gymnázium, to grade 9. The value added of multi-year gymnáziums and of basic schools did not differ in two of the three subjects — no significant difference was found in mathematics and language skills. Only in reading literacy did the gymnáziums show slightly higher value added (Greger, D. et al., 2017; 2022, in Nilsen, T. et al. (Eds.), International Handbook of Comparative Large-Scale Studies in Education, Springer).

Straková (2010, Sociologický časopis), using data from PISA 2000, PISA 2006, and the longitudinal PISA-L, found that the average value added of gymnáziums was actually lower than that of secondary vocational schools, with large differences among individual gymnáziums. The absolute performance advantage of gymnázium students is therefore attributable predominantly to a selection effect — the admission of capable pupils from families of high socioeconomic status.

PISA data consistently rank the Czech Republic among the countries with the highest between-school variance in the OECD — approximately 50% of the differences in results are explained at the school level, compared with roughly 30% on average across the OECD. The raw differences between gymnázium students and basic-school pupils in PISA are typically 80–120 points, but this reflects the entry quality of the pupils, not the transformative effect of education.

The work of Daniel Münich and colleagues from CERGE-EI and the IDEA analytical group has documented systemic problems: an extensive "preparation industry" for gymnázium entrance exams with a pronounced socioeconomic bias (Federičová and Münich, 2014); the negative impacts of "siphoning off" capable pupils from basic schools; and the economic inefficiency of educational inequalities, estimated at 18 billion CZK per year (IDEA and PAQ Research, 2020).

Münich and colleagues (2022, IDEA 9/2022) further demonstrated that, at equal ability, pupils from more strictly grading schools have lower academic aspirations: 87% of ninth-graders with an A in mathematics want to go to university, compared with only 39% of those with a C. Moreover, teachers grade differently depending on gender and socioeconomic status — the system therefore misallocates talents.

This finding is crucial: the problem of the Czech system does not lie in the principle of academic education (the German evidence clearly shows that a high-quality academic track supports cognitive development), but in the fact that Czech multi-year gymnáziums function primarily as a selection mechanism, not a development mechanism.

The strongest Central European quasi-experimental evidence comes from the Polish reform of 1999, which replaced the two-track system (8 years of basic school + 3–5 years of differentiated secondary education) with a three-stage system featuring a three-year comprehensive lower-secondary school (gimnazjum), thereby effectively postponing sorting by one year.

Jakubowski, Patrinos, Porta, and Wiśniewski (2016, Education Economics), using matching and difference-in-differences methods on PISA data, demonstrated that for students who under the old system would have entered basic vocational education, the reform improved results by approximately one standard deviation — an effect equivalent to roughly one year of instruction. For academically oriented pupils, the effect was negligible. Poland's PISA scores improved dramatically: mathematics from 470 points in 2000 to 495 in 2006, reading from 479 to 508 — from below the OECD average to above it.

Drucker, Horn, and Jakubowski (2022, Journal for Labour Market Research) subsequently demonstrated lasting economic consequences as well: the reform raised the probability of employment by approximately 3 percentage points and earnings by approximately 4%.

The Finnish comprehensive reform of the 1970s produced similar results: Pekkarinen and colleagues (2009, Journal of Public Economics), in a difference-in-differences design using military test data, showed that the reform reduced the intergenerational income elasticity from 0.30 to 0.23 — that is, it substantially improved social mobility, and did so without harming the results of advantaged students.

Hanushek and Woessmann (2006, Economic Journal), using international data, demonstrated that early tracking increases inequality without raising average performance in a compensating way. There is no trade-off between equality and quality — systems with late tracking (Finland, Canada, Korea) achieve better results on both dimensions. Strello and colleagues (2021) confirmed this across 21 cycles of the international PISA, PIRLS, and TIMSS tests from 75 countries over a twenty-year period.

The economics literature adds a further argument through an analysis of labor outcomes over the life cycle.

Hanushek, Schwerdt, Woessmann, and Zhang (2017, Journal of Human Resources) analyzed data from the International Adult Literacy Survey (IALS) covering 11 countries, supplemented by the German Mikrozensus and Austrian administrative data. Graduates of vocational education have higher initial employment, but this advantage reverses around age 50 — after that they have lower employment rates and lower earnings. This phenomenon is most visible in countries with a developed apprenticeship system (Denmark, Germany, Switzerland). As the mechanism, they identified differential access to lifelong learning: with increasing age, generally educated workers are significantly more often recipients of further professional training.

The speed at which specific skills become obsolete is illustrated by Deming and Noray (2020, Quarterly Journal of Economics): 29% of a given firm's job postings for the same position contained at least one new skill requirement within twelve years. Graduates of applied science and technical fields (STEM) earn 44% more than graduates of other fields at age 24, but this premium falls to just 14% by age 35.

Weber (2014, International Journal of Manpower), using Swiss data, demonstrated that education based on conceptual thinking protects workers against the depreciation of human capital more effectively than vocational education. Hanushek and Woessmann, in a series of papers (2008, 2012, 2015), demonstrated that the cognitive skills of the population — not the mere length of schooling — are the decisive factor in economic growth: raising test scores by one standard deviation increases annual GDP growth by approximately 1.3–2 percentage points over a forty-year period.

No honest analysis may overlook the counterarguments, and they do exist.

First, vocational education has real short-term advantages. Graduates of apprenticeships and secondary vocational schools enter the labor market faster and with lower unemployment in the early years of their careers. Silliman and Virtanen (2022, American Economic Journal: Applied Economics), in a Finnish regression-discontinuity design, even found a 7% earnings advantage for vocational education at age 31 with no sign of decline into the mid-thirties. In countries with a high-quality dual system (Finland, Switzerland), the trade-off may be less pronounced.

Second, Malamud and Pop-Eleches (2010, Review of Economics and Statistics), studying the Romanian reform of 1973, found that moving students from vocational into general education led to comparable levels of employment and earnings — suggesting that part of the differences between types of education is driven by a selection effect, not by the education itself.

Third, direct brain-imaging studies comparing the development of academic and vocational students do not yet exist. The claim about a neuroscientific mechanism is therefore based on indirect evidence and logical inference — not on direct observation.

Fourth, Korthals and colleagues (2022) pointed out that a higher educational track does improve cognitive outcomes, but worsens non-cognitive skills — self-concept and motivation. Students at the bottom of the academic track may suffer from comparison with more successful classmates.

Fifth, the Czech problem may lie not primarily in tracking itself, but in the quality of its implementation. If Czech gymnáziums actually produced value added comparable to the German ones, the overall balance would look different.

These objections are legitimate. They do not, however, overturn the fundamental findings: (1) education causally raises intelligence; (2) the academic track amplifies this effect; (3) early tracking increases inequalities without a compensating benefit to average performance; and (4) the adolescent brain is maximally malleable during the tracking period and should be stimulated as broadly as possible.

The convergence of evidence from neuroscience, psychology, and the economics of education points to a fundamental mismatch in timing. The decision about an educational track is made during a period when the frontal cortex — the seat of rational planning and weighing long-term consequences — is undergoing intensive remodeling. Steinberg's dual-systems model suggests that at this age the reward system prevails over rational planning. The decision is therefore made by a brain that is not biologically fully prepared for it.

Synaptic pruning in adolescence increases efficiency but reduces flexibility. Narrowly focused vocational education during the period of intensive pruning risks "preforming" the brain for a narrow repertoire of skills — precisely during the period of greatest potential for broad development. The maximal malleability of the association areas of the cortex around the age of fifteen should be devoted to varied cognitive stimulation, not to premature specialization.

The international evidence is consistent: systems with late tracking achieve better results in both equality and quality. The Polish reform unambiguously demonstrated that postponing tracking by a single year dramatically improves the results of the weakest students without harming the results of the strongest — with measurable benefits persisting into the labor market. The Czech system, with its multi-year gymnáziums that show minimal value added compared with basic schools, instead creates a structural Matthew effect: initial socioeconomic differences are amplified by a divergent educational environment.

What we do not know — and what would need to be researched — is direct imaging evidence of how the brain development of academic and vocational students differs. Such a study would require long-term magnetic resonance tracking of groups of pupils before and after the transition to secondary school, with control for selection effects. It would be costly, but its results could provide a decisive answer to the question of whether the type of education shapes the brain's architecture.

Until then, we have at our disposal a convergent, albeit indirect, chain of evidence: neuroscience shows that the adolescent brain is maximally malleable and formable; psychology and the economics of education demonstrate that academic education causally raises intelligence more than vocational education; and international comparisons show that late tracking is both fairer and more effective.

David Epstein, in his book Range (2019), formulated a principle that this evidence supports: in a complex and rapidly changing world, people with a broad range thrive, not narrow specialists. Delayed specialization leads to a better match between abilities and career path and to greater adaptability for an unpredictable future.

The brain of a fifteen-year-old is not a smaller version of the adult brain. It is a construction site in full operation — with cranes, scaffolding, and wet concrete. And on a construction site, it is better to build with a broad architectural plan than with a narrow specialization in a single type of room.

This article draws on peer-reviewed scientific studies, meta-analyses, and international datasets (PISA, PIAAC, IALS, ABCD). The key causal studies use natural experiments, instrumental variables, regression discontinuity, and propensity score matching to isolate causal effects from selection. The Czech evidence comes from the longitudinal CLoSE study, PISA data, and the work of the IDEA analytical group at CERGE-EI. The article acknowledges its limitations: a direct imaging comparison of educational tracks does not exist; the Finnish and Scandinavian results may not be fully transferable to the Central European context; Czech multi-year gymnáziums differ qualitatively from German Gymnasiums.

Information cutoff date: February 2026.

Data and sources were verified as of the date of processing, but the situation may change. Before using this for decision-making, we recommend updating the data and independently verifying the key findings.

Neuroscience of the adolescent brain:

Gogtay, N. et al. (2004). Dynamic mapping of human cortical development during childhood through early adulthood. PNAS, 101(21), 8174–8179.

Giedd, J. N. et al. (1999). Brain development during childhood and adolescence: a longitudinal MRI study. Nature Neuroscience, 2(10), 861–863.

Casey, B. J. et al. (2018). The Adolescent Brain Cognitive Development (ABCD) study. Developmental Cognitive Neuroscience, 32, 1–144.

Petanjek, Z. et al. (2011). Extraordinary neoteny of synaptic spines in the human prefrontal cortex. PNAS, 108(32), 13281–13286.

Averbeck, B. B. (2022). Pruning recurrent neural networks replicates adolescent changes in working memory and reinforcement learning. PNAS, 119(22), e2121331119.

Miller, D. J. et al. (2012). Prolonged myelination in human neocortical evolution. PNAS, 109(41), 16480–16485.

Steinberg, L. (2008). A social neuroscience perspective on adolescent risk-taking. Developmental Review, 28(1), 78–106.

Galván, A. et al. (2006). Earlier development of the accumbens relative to orbitofrontal cortex. Journal of Neuroscience, 26(25), 6885–6892.

Reynolds, L. M. et al. (2018). Mesocorticolimbic dopamine pathways across adolescence. Cerebral Cortex, 29(3), 1116–1127.

Sydnor, V. J. et al. (2023). Intrinsic activity development unfolds along a sensorimotor–association cortical axis in youth. Nature Neuroscience, 26, 638–649.

Fuhrmann, D., Knoll, L. J. & Blakemore, S.-J. (2015). Adolescence as a sensitive period of brain development. Trends in Cognitive Sciences, 19(10), 558–566.

Dahl, R. E. et al. (2018). Importance of investing in adolescence from a developmental science perspective. Nature, 554, 441–450.

Konrad, K., Firk, C. & Uhlhaas, P. J. (2013). Brain development during adolescence. Deutsches Ärzteblatt International, 110(25), 425–431.

Baker, S. T. et al. (2025). The connecting brain in context. Developmental Cognitive Neuroscience, 71, 101486.

Critical and sensitive periods:

Hartshorne, J. K., Tenenbaum, J. B. & Pinker, S. (2018). A critical period for second language acquisition. Cognition, 177, 263–277.

Tervo-Clemmens, B. et al. (2023). A canonical trajectory of executive function maturation. Nature Communications, 14, 6922.

Blakemore, S.-J. & Mills, K. L. (2014). Is adolescence a sensitive period for sociocultural processing? Annual Review of Psychology, 65, 187–207.

The causal effect of education on intelligence:

Ceci, S. J. (1991). How much does schooling influence general intelligence and its cognitive components? Developmental Psychology, 27(5), 703–722.

Brinch, C. N. & Galloway, T. A. (2012). Schooling in adolescence raises IQ scores. PNAS, 109(2), 425–430.

Ritchie, S. J. & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358–1369.

Cliffordson, C. & Gustafsson, J.-E. (2008). Effects of age and schooling on intellectual performance. Intelligence, 36(2), 143–152.

Carlsson, M., Dahl, G. B., Öckert, B. & Rooth, D.-O. (2015). The effect of schooling on cognitive skills. Review of Economics and Statistics, 97(3), 533–547.

Ritchie, S. J., Bates, T. C. & Deary, I. J. (2015). Is education associated with improvements in general cognitive ability, or in specific skills? Developmental Psychology, 51(5), 573–582.

Tracking and cognitive development:

Becker, M. et al. (2012). The differential effects of school tracking on psychometric intelligence. Journal of Educational Psychology, 104(3), 682–699.

Guill, K., Lüdtke, O. & Köller, O. (2017). Academic tracking is related to gains in students' intelligence over four years. Learning and Instruction, 47, 43–52.

Traini, C., Kleinert, C. & Bittmann, F. (2021). How does exposure to a different learning environment affect student achievement? Research in Social Stratification and Mobility, 76, 100625.

Herrmann, J. & Bach, M. (2025). Ability grouping in German secondary schools. British Journal of Educational Psychology, 95(2), 578–602.

Matthewes, S. H. (2021). Better together? Heterogeneous effects of tracking on student achievement. The Economic Journal, 131(635), 1269–1307.

Korthals, R. A., Schils, T. & Borghans, L. (2022). Track placement and the development of cognitive and non-cognitive skills. Education Economics, 30(5), 540–559.

Pekkala Kerr, S., Pekkarinen, T. & Uusitalo, R. (2013). School tracking and development of cognitive skills. Journal of Labor Economics, 31(3), 577–602.

Tracking and inequality:

Hanushek, E. A. & Woessmann, L. (2006). Does educational tracking affect performance and inequality? Economic Journal, 116(510), C63–C76.

Strello, A. et al. (2021). Does tracking affect the relationship between socioeconomic background and achievement? Evidence from 20 years of international assessments. British Journal of Sociology of Education, 42(5–6), 888–911.

Borghans, L. et al. (2020). What can we learn from tracking at early ages? IZA Discussion Paper, No. 13048.

Czech and Central European context:

Greger, D. et al. (2017; 2022). CLoSE longitudinal study. In International Handbook of Comparative Large-Scale Studies in Education, Springer.

Straková, J. (2010). The value added of gymnáziums. Sociologický časopis, 46(2), 187–210.

Münich, D. et al. (2022). Grading and educational aspirations. IDEA Study 9/2022, CERGE-EI.

Federičová, M. & Münich, D. (2014). Factors influencing pupils' transition to the eight-year gymnázium. CERGE-EI.

IDEA & PAQ Research (2020). Educational inequalities and economic inefficiency.

Jakubowski, M. et al. (2016). The impact of the 1999 education reform in Poland. Education Economics, 24(6), 557–572.

Drucker, L., Horn, D. & Jakubowski, M. (2022). The labour market effects of the Polish educational reform. Journal for Labour Market Research, 56, Article 13.

Pekkarinen, T., Uusitalo, R. & Kerr, S. (2009). School tracking and intergenerational income mobility: Evidence from the Finnish comprehensive school reform. Journal of Public Economics, 93(7-8), 965–973.

Horn, D. (2013). Diverging performances. Research in Social Stratification and Mobility, 32, 25–43.

Federičová, M. (2021). Tracking and peer effects. CERGE-EI Working Paper 709.

Neuroscience of learning and cognitive reserve:

Zatorre, R. J., Fields, R. D. & Johansen-Berg, H. (2012). Plasticity in gray and white. Nature Neuroscience, 15(4), 528–536.

Stern, Y. (2009). Cognitive reserve. Neuropsychologia, 47(10), 2015–2028.

Diamond, A. & Lee, K. (2011). Interventions shown to aid executive function development. Science, 333(6045), 959–964.

Bialystok, E., Craik, F. I. M. & Luk, G. (2012). Bilingualism: consequences for mind and brain. Trends in Cognitive Sciences, 16(4), 240–250.

Economic consequences of education:

Hanushek, E. A., Schwerdt, G., Woessmann, L. & Zhang, L. (2017). General education, vocational education, and labor-market outcomes over the lifecycle. Journal of Human Resources, 52(1), 48–87.

Deming, D. J. & Noray, K. (2020). Earnings dynamics, changing job skills, and STEM careers. Quarterly Journal of Economics, 135(4), 1965–2005.

Weber, S. (2014). Human capital depreciation and education level. International Journal of Manpower, 35(5), 613–642.

Hanushek, E. A. & Woessmann, L. (2012). Do better schools lead to more growth? Journal of Economic Growth, 17(4), 267–321.

Silliman, M. & Virtanen, H. (2022). Labor market returns to vocational secondary education. American Economic Journal: Applied Economics, 14(1), 197–224.

Malamud, O. & Pop-Eleches, C. (2010). General education versus vocational training. Review of Economics and Statistics, 92(1), 43–60.

Hampf, F. & Woessmann, L. (2017). Vocational vs. general education and employment over the life-cycle. CESifo Economic Studies, 63(3), 255–269.

Epstein, D. (2019). Range: Why Generalists Triumph in a Specialized World. Riverhead Books.

Creation transparency

The concept, structure, and editorial line of the article are the work of the author, who prepared the content outline, established the key theses, and directed the entire creative process. Generative AI (Claude Opus 4.6, Anthropic) was used as a tool for research, fact-checking, and elaborating the author's draft.

The author verified the key findings and approved the final wording. No part of the text was published without conscious authorial control. The factual data were verified against the publicly available sources cited in the text.

This procedure complies with the transparency principles of EU Regulation 2024/1689 (AI Act). #poweredByAI