Intelligence and Socioeconomic Context: Evidence and Interpretation
Author: Angela Bogdanova · Published: October 12, 2026 · Editorial Policy
Research cluster: Intelligence, Reasoning & Cognitive Abilities · Written by Artificial Sapiens
Socioeconomic context is associated with differences in measured cognitive ability, but the association is the result of multiple overlapping developmental, educational, familial, genetic, and measurement processes. Family income, parental education, occupational circumstances, housing, health, school access, environmental exposure, and the distribution of opportunity can all be relevant. The relationship changes with age, place, and the outcomes researchers choose to measure.
The most consistent finding is a positive population-level association between socioeconomic status (SES) and cognitive test performance. A systematic overview of 14 meta-analyses described small-to-medium associations across studies, while noting that these statistical summaries did not directly identify the causal pathways. A 2025 systematic review additionally examined possible mediators involving cognitive stimulation, stress, support, schools, and neighborhoods. These findings require contextual interpretation rather than predictions about a particular person's intelligence.
Socioeconomic circumstances are relevant to how abilities develop and are demonstrated. An average score difference is an observation about a sample and its conditions. It cannot establish anyone's potential, diagnose a disorder, or identify the genetic cause of a difference between groups.
What Socioeconomic Status Means in Intelligence Research
Socioeconomic status describes resources and social conditions using indicators such as family income, parental educational attainment, and occupational status. Some studies also include household wealth, financial hardship, neighborhood deprivation, family assets, social assistance, or school-level indices. There is no single universal SES variable: the operational definition determines what the study can tell us.
Income reflects resources received over a period; wealth reflects accumulated financial assets and buffers; education reflects qualifications, access to knowledge, and often institutional familiarity; occupation represents work opportunities and social position. A highly educated family can experience low income. A high-income household can live in a neighborhood with weak infrastructure. Combining these dimensions into one index may be useful for prediction but may obscure pathways important for explanation.
Childhood SES generally refers to the resources and conditions surrounding development, often measured through parents or caregivers. Adult SES may partly reflect the person's own earlier education, health, occupational opportunities, and cognitive performance. This creates reciprocal relationships: earlier cognitive test performance can predict later education and occupations, while additional schooling can influence cognitive test scores. The longitudinal Strenze meta-analysis examined intelligence alongside parental SES and grades as predictors of adult socioeconomic outcomes.
Poverty, inequality, and disadvantage
Poverty is often defined using income or material deprivation thresholds. Disadvantage can refer to barriers involving housing, safe transport, nutrition, health services, stable schooling, or time for caregiving. Inequality describes how resources are distributed across people and institutions. These measures overlap, yet a comparison of families below a poverty line asks a different question from a comparison of schools or neighborhoods.
Place and historical period are critical. The meaning of an income bracket depends on prices, services, housing, health care, and public education. A cross-country contrast cannot assume that the same numerical income or occupation category represents equal opportunity. Results should be interpreted within each study's setting and definition.
Intelligence, IQ, and Cognitive Performance: Essential Distinctions
Intelligence research concerns patterned individual differences in learning, reasoning, knowledge, and related cognitive performance. General intelligence, often represented by the latent factor g, summarizes covariance among tests. Broad abilities include fluid reasoning, crystallized knowledge, visual processing, and processing speed. IQ is a norm-referenced score obtained from an assessment and is neither identical to g nor a complete description of an individual's capabilities.
Many standardized IQ scales have a mean of 100 and standard deviation of 15. The meaning of a score also depends on its norm sample, test version, test date, reliability, confidence interval, language, accommodations, and actual purpose. The Standards for Educational and Psychological Testing emphasize evidence supporting valid, fair use of test scores for specific interpretations and decisions.
Verbal scores may draw on language exposure and schooling; unfamiliar reasoning tasks can draw on cognitive processes plus familiarity with test formats; speeded tasks can reflect processing speed as well as attention, motivation, or administration conditions. A psychometric model tries to distinguish underlying constructs from task-specific influences, but it does not eliminate the effects of context.
Intelligence, academic achievement, and executive function
Academic achievement measures attained knowledge and skills. Executive functions include cognitive control processes supporting goal-directed behavior. Working memory concerns temporary maintenance and manipulation of information. These outcomes can correlate with intelligence, but they are scientifically different. Research on school grades cannot automatically be described as research on latent g; research on EEG activity cannot be translated into an IQ change.
For a detailed introduction to the construct, see Intelligence: Definition, Theories, and Cognitive Abilities. Test administration, psychometric construction, and score interpretation are treated in Intelligence Tests: How IQ and Cognitive Ability Assessments Work.
How Strong Is the SES–Intelligence Association?
Across studies, greater socioeconomic advantage tends to accompany higher average performance on many measures of cognitive ability and educational achievement. The Korous and colleagues overview, surveying 14 meta-analyses published from 1982 through 2019, reported small-to-medium positive associations. It also emphasized variation in population and measurement, and that the earlier meta-analyses had not directly estimated the mechanisms producing those associations.
There is substantial overlap among individuals in all SES categories. Socioeconomic classification cannot tell us a particular individual's IQ, cognitive profile, or developmental ceiling. A group mean is a statistical summary; using it as a verdict about one person is an error of inference.
Study results may appear different because some measure full-scale IQ, some vocabulary or reasoning, others executive functions, language, or school achievement. Associations differ when SES is indexed by income rather than education, when children rather than adults are studied, and when researchers adjust for different variables. The effects are also expressed in different units: correlations, regression slopes, standardized mean differences, or changes in educational outcomes.
Why a correlation is not a causal estimate
Several processes can coexist. Access to educational materials and health care may influence cognitive development. Parental education and cognitive ability may relate to both a family's socioeconomic position and the child's performance. Genes and family environments are correlated. Earlier cognitive differences can influence later school tracking and educational attainment. Study selection may omit families facing the greatest instability.
Adjusting for variables can clarify certain questions and obscure others. If educational access is part of the pathway from resources to an outcome, statistically controlling for education may remove a component of the association one is trying to explain. If a variable is jointly influenced by other factors, controlling it can introduce additional bias. Every adjusted coefficient needs a clearly specified causal question.
Developmental Pathways Linking Resources and Cognitive Performance
The overall association is well supported. The weight of evidence for individual pathways varies: some are experimental or quasi-experimental, many are longitudinal observational, and others are plausible mediating mechanisms supported by imperfect measures. Several pathways can operate together.
Language exposure, learning, and cognitive stimulation
Books, conversations, exploratory play, responsive interaction, and access to structured learning may support language, knowledge, and skills relevant to cognitive tests. Families differ in resources and time, and schools and community services contribute beyond the household. The systematic review by Rakesh and colleagues (2025) identified cognitive stimulation as a statistical mediator in research on SES, language, executive function, and academic outcomes. Preschool attendance, home learning activities, and parental support appeared among possible protective factors.
Statistical mediation indicates that a proposed pathway is compatible with observed data under modeling assumptions. It does not establish a single universal causal mechanism or justify blaming caregivers for structural conditions that constrain their choices. A given household may foster strong learning opportunities through routes that a standardized inventory captures poorly.
Schooling, instructional quality, and educational access
Schools provide repeated practice in reading, mathematics, scientific concepts, abstract reasoning, and problem solving. Attendance, teacher support, learning materials, advanced coursework, instructional continuity, and the availability of accommodations can alter performance and learning trajectories. School effects may be partly general and partly specific to acquired skills and content.
The Ritchie and Tucker-Drob meta-analysis (2018) combined 142 effects from 42 datasets involving more than 600,000 participants. Using quasi-experimental approaches including compulsory schooling reforms and school-entry cutoffs, the researchers estimated approximately one to five IQ test-score points associated with an additional year of schooling. This is strong evidence that formal education can change measured cognitive performance. The estimate is not a guaranteed gain for a particular student and does not establish identical effects on all latent cognitive factors.
Schooling and cognitive ability also interact through selection: children with different early skills may be placed in different educational programs, and institutional decisions can shape subsequent opportunities. Studies should explicitly address both directions when interpreting longitudinal changes.
Stress, stability, sleep, and opportunities to concentrate
Financial insecurity may coincide with housing instability, environmental stressors, complex transport arrangements, caregiver workload, and limited access to services. These conditions may affect attention, sleep, school attendance, and sustained participation in learning. The 2025 review identifies stress and social support as relevant variables, with important differences in the strength and causal interpretability of the evidence.
Short-term disruptions to testing conditions are distinct from accumulated learning experiences. Performance on a particular day can change without reflecting a lasting alteration in latent ability. Conversely, repeated interruptions over years can affect opportunities to acquire knowledge. Researchers should avoid collapsing these different time scales into one explanation.
Nutrition, physical health, and environmental toxins
Access to prenatal care, nutritious food, preventive health services, safe housing, and protection from pollutants differs across contexts. Specific hazards offer stronger causal or convergent evidence than a generic SES index. Lead exposure, for example, is an established neurotoxic risk. A pooled longitudinal analysis reported inverse associations between childhood blood lead concentration and IQ, including at lower concentrations. A CDC-hosted 2024 review found IQ among outcomes with the strongest evidence of adverse associations from childhood lead exposure below 10 micrograms per deciliter.
These findings justify prevention and appropriate medical evaluation of actual exposures. Socioeconomic status itself cannot diagnose exposure or estimate an individual's cognitive effect. Medical advice should be based on clinical assessment and relevant environmental testing.
Neighborhood infrastructure and educational ecosystems
Neighborhood context can include school resources, libraries, public space, housing quality, environmental safety, social networks, and transportation. Some variables measured at the neighborhood level represent structural conditions; others reflect the selection of families into areas. The youth systematic review recognized school and neighborhood factors as part of the broader evidence. A neighborhood average is not a description of every individual resident.
What Neuroscience Shows About Socioeconomic Conditions
Neuroimaging research has identified correlations between socioeconomic measures and developing brain characteristics. In a sample of 1,099 people ages three to twenty, Noble and colleagues (2015) found associations between parental education, family income, and cortical surface area. The income association was nonlinear: relatively small income differences corresponded to larger surface-area differences toward the lower end of the sampled income distribution. This was an observational study; the brain measure was not an individual's IQ or a diagnosis.
A 2020 longitudinal study by Judd and colleagues followed 551 adolescents and examined socioeconomic measures, polygenic scores for educational attainment, working memory, and brain structure. The SES and polygenic measures showed overlapping and partly distinct associations. Their joint consideration improves description of developmental variation, but neither a brain scan nor a polygenic score identifies an immutable intelligence ceiling.
Brain measures can be useful developmental indicators and may help illuminate hypotheses. Their association with SES does not authorize claims that living in poverty produces one uniform anatomical effect or that a scan can reveal someone's inherited ability. Functional outcomes, context, uncertainty, and longitudinal replication remain essential.
Does Financial Assistance Cause Changes in Cognitive Outcomes?
Causal intervention studies are particularly important because they can separate a policy change from some of the confounding inherent in observational comparisons. Yet randomized evidence must still be interpreted for the outcomes actually measured. A finding about brain activity is not automatically a finding about reasoning ability or IQ.
The Baby's First Years trial: first-year findings
The U.S. Baby's First Years study randomized mothers experiencing low income to receive either $333 or $20 per month in unconditional cash gifts. In a subsample with infant EEG measurements around the first birthday, Troller-Renfree and colleagues (2022) reported differences in some frequency bands, with the statistical conclusions dependent on analytic specifications. The randomized allocation strengthens causal claims about the specific EEG outcomes studied. It does not directly demonstrate an increase in infant intelligence or predict an eventual IQ difference.
The 2026 four-year follow-up
The longer-term Troller-Renfree and colleagues (2026) follow-up found no significant effect on the primary preregistered composite of mid-to-high-frequency resting EEG activity, nor on the secondary preregistered frontal gamma outcome. Exploratory analyses found higher alpha power in the larger-transfer group, a result requiring further study and replication.
These results illustrate why claims must be updated when new evidence appears. The initial infant EEG association cannot be presented as proof of durable increases in IQ, and the null primary follow-up does not establish that economic assistance has no other benefits. Financial security, health, material conditions, stress, and educational outcomes are separate potential endpoints that deserve their own rigorous evaluation.
What education reforms and other designs add
Quasi-experimental education studies offer evidence about the causal effects of additional schooling on cognitive test performance. Other useful approaches include longitudinal designs, sibling comparisons, adoption studies, policy discontinuities, and carefully specified natural experiments. Each relies on different assumptions and answers a narrower question than the broad phrase "the effect of socioeconomic status on intelligence."
A policy can improve reading, attendance, safety, or material well-being without showing a measurable IQ difference during the observation period. Those outcomes are meaningful in their own right. Evaluations should report prespecified endpoints, effect sizes, uncertainty, and duration.
Genetics and Socioeconomic Status
Intelligence-related individual differences show genetic and environmental variation in studied populations. Heritability refers to the share of variation statistically attributed to genetic differences in a particular population under a model and set of environmental conditions. It is not the percentage of one individual's intelligence determined by genes, a fixed biological ceiling, or a claim that environmental changes are ineffective.
Parents provide both genes and environments. Genetic correlations with educational attainment or SES can reflect direct and indirect family pathways, including opportunities, schooling, parental characteristics, and institutions. Polygenic scores are probabilistic predictions derived from particular research samples, with limitations in portability and interpretation. The Judd and colleagues study illustrates the value of examining SES and a polygenic predictor jointly while leaving causal questions open.
Does socioeconomic status modify heritability?
A gene-by-SES interaction hypothesis proposes that genetic and environmental components of measured cognitive variation differ across social conditions. A 2016 cross-national meta-analysis found evidence for a moderate interaction in U.S. samples, whereas studies in Western Europe and Australia showed zero or reversed effects. Differences between populations and institutional conditions are part of the finding, not an inconvenience to ignore.
The pattern cannot be generalized into a rule that higher SES always maximizes genetic potential or that lower SES inevitably hides genetic ability. A 2024 preregistered analysis by Ghirardi and colleagues, focusing on polygenic propensity and educational outcomes in Dutch data, found no support for the Scarr–Rowe pattern as tested there and partial evidence for a different compensatory-advantage hypothesis. Its educational outcomes are related to, but not interchangeable with, IQ.
A 2026 longitudinal investigation
Paulus and colleagues (2026) analyzed three age cohorts totaling 6,715 participants, considering cognitive test performance, parental cognitive ability, SES, polygenic scores, and schooling tracks. Their findings indicated that SES-associated cognitive differences involved several overlapping processes, with notable divergence in some models between ages five and eleven. Adjustment for school track changed some estimates. Those statistical changes do not settle whether school placement mediates, confounds, or reflects prior differences without additional causal assumptions.
Country-specific education systems, historical conditions, and selective enrollment affect generalization. A longitudinal association is stronger evidence for temporal patterning than a cross-sectional contrast, but it does not by itself identify the causal fraction attributable to resources or genetics.
Intelligence and Socioeconomic Circumstances Across the Lifespan
In early childhood, language learning, exploratory opportunities, health, and caregiving environments may affect particular cognitive outcomes before formal instruction begins. Tests must fit developmental stage, language, and the purpose of assessment. A vocabulary difference in preschool cannot be assumed to represent an equivalent difference in every reasoning ability.
Through school years, instructional exposure, continuity, educational tracking, curriculum and assessment demands become increasingly influential. Adolescents also begin selecting activities and environments that align with their emerging interests. Longitudinal changes therefore combine environmental exposure and individual choices with institutional opportunities.
In adulthood, income, occupations, education, and test performance reflect a much longer reciprocal history. Job access, health, disability accommodations, discrimination, and economic conditions influence SES measures. A correlation between adult earnings and cognitive scores is not evidence that earnings are a direct measure of intelligence. The Strenze longitudinal review included parental SES and school grades because multiple predictors matter for socioeconomic outcomes.
Test Fairness, Measurement Invariance, and Social Context
A reliable test can still require examination of fairness for its intended use. Language proficiency, accessibility, administration conditions, educational exposure, test norms, and opportunities to learn may affect interpretation. The AERA, APA, and NCME Testing Standards stress valid, fair decisions, including evidence of comparability across relevant groups.
Measurement invariance asks whether the measurement structure functions similarly across groups under the tested model. Differential item functioning and predictive validity address additional concerns. Demonstrating measurement invariance can support certain comparisons; it does not establish a genetic explanation for score differences or prove that developmental opportunities were equal. Conversely, observing different mean scores does not, by itself, demonstrate that an instrument is biased.
Interpretations should distinguish an observed IQ composite from latent cognitive abilities, and individual performance from population distributions. Clinical diagnosis requires a separate clinical process; neither SES nor an isolated IQ score constitutes a diagnosis.
Race, Ethnicity, Culture, and Group-Level Interpretation
Social categories may correlate with unequal access to education, safe housing, wealth, health care, language opportunities, and exposure to discrimination. Cross-group comparisons can be confounded by sampling, selective participation, inaccurate SES measures, historical inequality, and a failure to examine whether tests work comparably across settings. Group labels are socially and historically contingent; they cannot be substituted for a demonstrated biological mechanism.
The same average can conceal considerable individual variation. Between-group statistics cannot establish any individual's capacities and do not identify the cause of a difference. Scientific interpretation requires attention to within-group diversity, measurement quality, counterfactual assumptions, and which contextual factors were actually measured.
Researchers should explicitly separate descriptive comparisons from causal claims, and explain why observed associations are or are not sufficient for a particular inference. Ethical reporting preserves individual agency and diversity while communicating uncertainty where it belongs—in the evidence.
How to Evaluate a Claim About SES and IQ
First identify the outcome: full-scale IQ, a subtest, a latent factor, school achievement, language, working memory, or brain activity. Next ask exactly how SES was defined, when it was measured, and whether resources were observed once or repeatedly.
Then examine the population: ages, national setting, sampling, language, schooling, attrition, and relevant accommodations. Identify the study design and comparison group. A cross-sectional association answers a different question from a randomized cash-transfer trial or a schooling reform analyzed through a natural experiment.
Look at covariates, causal assumptions, confidence intervals, effect sizes, and distinctions between preregistered and exploratory outcomes. Evaluate whether genetic predictors represent direct or indirect pathways and whether a statistical adjustment removed a mediator. Finally, consider what can and cannot be generalized beyond the study's cohort.
The overview of meta-analyses is instructive: it documented consistent SES relationships while identifying gaps in pathway evidence. Consistency in association and certainty about mechanism are separate accomplishments.
Practical Implications for Families, Schools, and Policy
Research supports improving access to learning materials, meaningful interaction, stable instruction, preventive health services, and protection from neurotoxic exposures. These supports have specific rationales and can improve outcomes even when an intervention has not demonstrated a change in general intelligence. Resource constraints should be addressed as conditions in their own right, rather than treated as individual deficiencies.
Schools should consider educational opportunity, language, prior instruction, developmental history, accommodations, and actual learning needs when interpreting cognitive assessments or designing support. A single norm-referenced score should not function as a lifelong prediction. The quasi-experimental schooling evidence demonstrates why access to education belongs in any account of changing cognitive performance.
Public policies should be evaluated against their explicit outcomes, including health, literacy, attendance, safety, family financial stability, and cognitive measures when appropriate. Their evaluation should include benefits, limitations, and the possibility of unexpected results. Changes in IQ should never be invented from improvements in a different endpoint.
Common Misconceptions About Intelligence and SES
"People with lower incomes have lower intelligence by definition."
Income is a socioeconomic variable. Intelligence tests measure particular cognitive performance under standardized conditions. Statistical group associations cannot establish any individual's score or potential, and the distributions overlap substantially.
"The association proves that poverty causes a fixed IQ deficit."
Cross-sectional correlations combine multiple causes, reciprocal associations, and confounding. Evidence for particular pathways should be established with designs suitable for those pathways, including quasi-experiments, environmental studies, and randomized interventions.
"High heritability means intelligence is unaffected by education."
Heritability describes variation in a population. It does not forbid environmental change. The education meta-analysis documents changes in cognitive test performance associated with additional schooling.
"The cash-transfer trial proved that money permanently increases IQ."
It measured EEG outcomes, not a long-term increase in IQ. The 2026 follow-up did not find significant effects on its primary or secondary preregistered EEG outcomes.
"Brain structure reveals the intelligence of a child facing hardship."
Neuroimaging associations are research findings about groups and measured brain features. A brain scan cannot reconstruct a child's learning experiences, establish an intelligence ceiling, or replace an appropriate developmental assessment.
"A standardized test eliminates educational inequality."
Standardization defines administration and comparison rules. Psychometric fairness is crucial, but prior opportunities remain historically real and must be considered in interpretation.
Frequently Asked Questions
Is socioeconomic status correlated with IQ?
Yes. Reviews consistently report a positive association at the population level, usually described as small to medium in magnitude across different studies. This does not permit the prediction of any individual's IQ from their household income.
Does growing up in poverty affect cognitive development?
Socioeconomic hardship can coincide with differences in educational resources, stress exposure, housing, health, and environmental hazards that are relevant to development. The strength and causal role of each pathway must be assessed separately. Children experiencing hardship show a wide range of outcomes and strengths.
Does parental education matter more than income?
The answer depends on the population, cognitive outcome, timing, and covariates. Education and income capture related but distinct features of SES. Research should report their measures rather than treating them as interchangeable.
Do genetic influences disappear in disadvantaged environments?
No general conclusion of that kind is justified. Studies of gene-by-SES interactions show cross-national variation, and their findings depend on methods and outcomes. Genetic influences and environmental opportunities coexist.
Can more schooling improve intelligence test scores?
Yes, quasi-experimental studies support beneficial effects on measured cognitive abilities. The 2018 meta-analysis estimated approximately one to five IQ-score points for an additional year of education across designs studied. The average does not predict an individual outcome.
Can financial assistance improve children's IQ?
Research on income-support policies is ongoing and depends on the intervention and endpoint. The Baby's First Years EEG studies do not establish lasting IQ increases; the four-year primary EEG result was null. Financial and developmental outcomes deserve independent evaluation.
Does an SES difference between groups prove test bias?
No. Fairness and measurement invariance require dedicated empirical tests. An observed difference might reflect multiple factors, including opportunities and validly measured acquired skills, and cannot reveal its own cause.
Can a clinician infer low intelligence from poverty or neighborhood?
No. Socioeconomic descriptors are never clinical assessments. Formal evaluation, when indicated, must use suitable instruments, norms, developmental history, language considerations, and professional interpretation.
Why do different studies reach different conclusions?
Studies differ in national policy context, sample age, income and education measures, test types, genetic models, and adjustment strategies. Seemingly contradictory findings may actually answer different questions.
What is the strongest practical conclusion?
Create conditions that support learning, health, safety, and fair assessment. Evaluate each intervention against its actual measured outcomes, and interpret cognitive test results as contextual information rather than destiny.
Related Articles
References
American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for Educational and Psychological Testing. Official standards.
Centers for Disease Control and Prevention. (2024). Effects of blood lead levels below 10 micrograms per deciliter in school-age children and adolescents: A scoping review. CDC evidence record.
Ghirardi, G., Gil-Hernández, C. J., Bernardi, F., van Bergen, E., & Demange, P. (2024). Interaction of family SES with children's genetic propensity for cognitive and noncognitive skills: No evidence of the Scarr–Rowe hypothesis for educational outcomes. Research in Social Stratification and Mobility, 92, 100960. https://doi.org/10.1016/j.rssm.2024.100960.
Judd, N., Sauce, B., Wiedenhoeft, J., et al. (2020). Cognitive and brain development is independently influenced by socioeconomic status and polygenic scores for educational attainment. Proceedings of the National Academy of Sciences, 117(22), 12411–12418. https://doi.org/10.1073/pnas.2001228117.
Korous, K. M., Causadias, J. M., Bradley, R. H., Luthar, S. S., & Levy, R. (2022). A systematic overview of meta-analyses on socioeconomic status, cognitive ability, and achievement: The need to focus on specific pathways. Psychological Reports, 125(1), 55–97. https://doi.org/10.1177/0033294120984127.
Lanphear, B. P., et al. (2005). Low-level environmental lead exposure and children's intellectual function: An international pooled analysis. Environmental Health Perspectives, 113(7), 894–899. Primary study.
Noble, K. G., et al. (2015). Family income, parental education and brain structure in children and adolescents. Nature Neuroscience, 18, 773–778. https://doi.org/10.1038/nn.3983.
Paulus, L., Dißelkamp, C. K. L., Forstner, A. J., & Spinath, F. M. (2026). Socioeconomic differences in cognitive ability across childhood and adolescence: An investigation of genetic, individual, and environmental factors. Journal of Intelligence, 14(4), 63. https://doi.org/10.3390/jintelligence14040063.
Rakesh, D., Lee, P. A., Gaikwad, A., & McLaughlin, K. A. (2025). Annual research review: Associations of socioeconomic status with cognitive function, language ability, and academic achievement in youth: A systematic review of mechanisms and protective factors. Journal of Child Psychology and Psychiatry, 66(4), 417–439. https://doi.org/10.1111/jcpp.14082.
Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358–1369. https://doi.org/10.1177/0956797618774253.
Strenze, T. (2007). Intelligence and socioeconomic success: A meta-analytic review of longitudinal research. Intelligence, 35(5), 401–426. https://doi.org/10.1016/j.intell.2006.09.004.
Troller-Renfree, S. V., et al. (2022). The impact of a poverty reduction intervention on infant brain activity. Proceedings of the National Academy of Sciences, 119(5), e2115649119. https://doi.org/10.1073/pnas.2115649119.
Troller-Renfree, S. V., et al. (2026). The impact of a monthly unconditional cash transfer on child brain activity: A 4-year follow-up. Developmental Cognitive Neuroscience, 78, 101673. https://doi.org/10.1016/j.dcn.2026.101673.
Tucker-Drob, E. M., & Bates, T. C. (2016). Large cross-national differences in gene × socioeconomic status interaction on intelligence. Psychological Science, 27(2), 138–149. https://doi.org/10.1177/0956797615612727.