How do culture-fair IQ tests differ from traditional IQ tests?

Culture-fair IQ tests focus on **nonverbal, abstract reasoning tasks** -- such as pattern completion, matrix reasoning, and figure classification -- designed to minimize the influence of language and culturally specific knowledge. Traditional IQ tests like the **WAIS-IV** include substantial verbal components (vocabulary definitions, comprehension questions, information recall) that inherently favor individuals educated in the test's language and cultural context. Research consistently shows that score gaps between cultural groups are ***smaller on culture-fair tests*** than on traditional verbally loaded measures, though they are not completely eliminated (Jensen, 1998).

Can culture-fair IQ tests completely remove all forms of bias?

No. While culture-fair tests significantly reduce bias related to **language proficiency** and **culturally specific knowledge**, they cannot eliminate all sources of unfairness. Abstract reasoning skills are themselves influenced by educational quality, exposure to puzzles and games, and cognitive stimulation during development. The **Flynn Effect** -- showing that scores on culture-fair tests like Raven's Matrices have risen dramatically over generations -- demonstrates that environmental factors shape performance even on nonverbal measures. The consensus term among researchers is now ***"culture-reduced"*** rather than "culture-free" (Sternberg, 2020).

Why is it important to use multiple types of assessments alongside culture-fair tests?

A single test measures a narrow slice of cognitive ability. Combining culture-fair instruments with **verbal IQ tests**, **achievement measures**, **behavioral observations**, and **interview data** provides a more complete picture. For example, a student might score 115 on Raven's Matrices but 95 on a verbally loaded test due to limited English proficiency -- the discrepancy itself is diagnostically valuable, revealing that cognitive ability is being masked by language barriers. The **APA's Standards for Educational and Psychological Testing** (2014) explicitly recommends against making consequential decisions based on any single test score.

How can educators use culture-fair IQ tests to support diverse students?

Culture-fair assessments are particularly valuable for **gifted identification** in diverse schools. The Naglieri and Ford (2003) study demonstrated that using the **NNAT** instead of traditional verbally loaded tests increased the identification of gifted African American and Hispanic students significantly. Practical steps for educators include: **(1)** adopting nonverbal screening as the first stage of gifted identification; **(2)** using local norms rather than national norms when student populations differ significantly from the norming sample; **(3)** supplementing test scores with teacher observations, portfolios, and parent input; and **(4)** re-testing students after they have had time to adjust to new educational environments.

What are some limitations of culture-fair IQ tests in real-world settings?

Key limitations include: **(1)** *Residual cultural influence* -- abstract reasoning itself is trained through education and cognitive stimulation, which vary across cultures; **(2)** *Narrower measurement* -- by excluding verbal and knowledge-based content, culture-fair tests assess only fluid intelligence, missing crystallized abilities that are relevant to many real-world tasks; **(3)** *Familiarity effects* -- individuals who have never encountered multiple-choice formats or timed tests may underperform regardless of ability; **(4)** *Ceiling effects* -- some culture-fair tests lack sufficient difficulty at the upper end to differentiate among highly gifted individuals; **(5)** *Normative limitations* -- many tests lack locally calibrated norms for non-Western populations, forcing reliance on inappropriate comparison groups.

How do socioeconomic factors impact the fairness of IQ testing?

Socioeconomic status (SES) affects IQ test performance through multiple pathways: **nutrition** (iodine deficiency alone can reduce IQ by 10-15 points), **prenatal and early childhood healthcare**, **educational quality and duration**, **cognitive stimulation at home** (number of books, parental verbal interaction), and **chronic stress** (which impairs prefrontal cortex development). A meta-analysis by Sirin (2005) found a correlation of **r = 0.29** between SES and academic achievement. Critically, these factors affect performance on *all* IQ tests, including culture-fair ones -- because they shape the development of the very cognitive abilities the tests measure. This is why researchers like ***James Flynn*** argue that IQ tests measure a combination of innate potential and environmental advantage.

Are there technological advancements improving culture-fair IQ testing?

Yes, several promising developments are underway. **Computerized adaptive testing (CAT)** tailors item difficulty in real time, improving measurement precision while potentially selecting culturally appropriate items from calibrated banks. **Dynamic assessment** measures learning potential rather than static knowledge, which may be inherently more equitable. **Game-based assessments** reduce test anxiety and unfamiliarity with formal testing formats. **Neuroscience-based measures** (EEG, fMRI) may eventually provide direct indices of neural efficiency, though equitable access to brain-scanning technology remains a major barrier. The most promising direction may be **multi-method assessment platforms** that combine nonverbal reasoning, dynamic tasks, and adaptive algorithms to triangulate cognitive ability from multiple angles.

Culture-Fair IQ Tests: Do They Really Reduce Bias?

Understanding Culture-Fair IQ Tests

The pursuit of culture-fair IQ tests has been one of the most consequential efforts in the history of intelligence assessment. Traditional IQ tests -- built predominantly in Western, English-speaking contexts -- have long faced criticism for containing cultural bias that disadvantages individuals from different linguistic, ethnic, or socioeconomic backgrounds. This bias is not merely a theoretical concern: it has real consequences for educational placement, clinical diagnosis, and opportunity.

Culture-fair tests attempt to measure fluid intelligence -- the capacity to reason, identify patterns, and solve novel problems -- without relying on vocabulary, general knowledge, or culturally specific content. The goal is elegant in its simplicity: strip away everything that depends on where and how you were raised, and measure the cognitive engine underneath.

But can any test truly achieve this? The history of culture-fair testing, from Raymond Cattell's pioneering work in the 1940s to modern computerized adaptive assessments, reveals both genuine progress and stubborn limitations.

"Intelligence is what you use when you don't know what to do."
-- Jean Piaget, developmental psychologist

This article explores the mechanisms behind culture-fair testing, the landmark instruments that define the field, the evidence for and against their effectiveness, and what fair testing means in an increasingly interconnected world. Along the way, you can take our full IQ test or try a quick IQ assessment to experience different testing formats firsthand.

The Origins of Culture-Fair Testing: Cattell and the CFIT

The modern concept of culture-fair intelligence testing owes its foundation to Raymond B. Cattell, a British-American psychologist who drew a crucial distinction between two types of intelligence in 1963:

Fluid intelligence (Gf) -- the ability to reason abstractly, recognize patterns, and solve novel problems independently of prior knowledge
Crystallized intelligence (Gc) -- accumulated knowledge, vocabulary, and skills acquired through education and experience

Cattell recognized that traditional IQ tests heavily weighted crystallized intelligence, which is inherently shaped by culture, language, and educational opportunity. His response was the Culture Fair Intelligence Test (CFIT), first published in 1949 and revised multiple times since.

How the CFIT Works

The CFIT presents test-takers with nonverbal, figural tasks across four subtests:

Series completion -- identifying the next figure in a logical sequence
Classification -- selecting which figure does not belong in a group
Matrices -- finding the missing element in a pattern grid
Conditions (topology) -- identifying which figure satisfies stated spatial conditions

No reading, writing, or verbal communication is required beyond understanding basic instructions. The test is available in three scales for different age groups and ability levels.

"The purpose of the Culture Fair test is to provide a measure of intelligence that is relatively free of cultural and educational influences."
-- Raymond B. Cattell, Theory of Fluid and Crystallized Intelligence (1963)

CFIT Scale Structure

Scale	Target Population	Number of Items	Administration Time
Scale 1	Ages 4-8, adults with intellectual disabilities	8 subtests (various)	~30 min
Scale 2	Ages 8-14, average adults	4 subtests, 46 items	~25 min
Scale 3	Ages 14+, above-average adults	4 subtests, 50 items	~25 min

Cattell's work established the theoretical framework that virtually all subsequent culture-fair instruments have built upon: test fluid intelligence through abstract, nonverbal tasks.

Raven's Progressive Matrices: The Gold Standard

If Cattell provided the theory, John C. Raven provided the instrument that would become the most widely used culture-fair test in the world. First published in 1938, Raven's Progressive Matrices (RPM) asks test-takers to identify the missing piece in a visual pattern -- a deceptively simple format that taps deeply into abstract reasoning.

The Three Versions of Raven's Matrices

Version	Target Audience	Items	Difficulty	Primary Use
Coloured Progressive Matrices (CPM)	Children ages 5-11, elderly, clinical populations	36 items in 3 sets	Easy to moderate	Screening, pediatric assessment
Standard Progressive Matrices (SPM)	Ages 8 to adult	60 items in 5 sets	Moderate	General population assessment
Advanced Progressive Matrices (APM)	Above-average adults, university students	48 items in 2 sets	Difficult	Gifted identification, research

Each problem presents a matrix (typically 3x3 or 2x2) of geometric designs with one cell empty. The test-taker must select the correct piece from six or eight options. Items progress from simple perceptual matching to complex multi-rule reasoning involving:

Pattern continuation
Quantitative progression
Figure addition and subtraction
Distribution of elements across rows and columns

Why Raven's Became the Global Standard

Raven's Matrices gained worldwide adoption for several reasons:

Minimal language requirements -- instructions can be demonstrated rather than verbally explained
Strong theoretical grounding -- it is one of the best single measures of Spearman's g factor (general intelligence)
Extensive cross-cultural norming -- data from over 100 countries
Simplicity of administration -- requires no specialized equipment

"Raven's Progressive Matrices is probably the purest measure of general intelligence that we have."
-- Arthur Jensen, The g Factor (1998)

Research by John Raven Jr. and colleagues (2000) compiled normative data from populations across Africa, Asia, Europe, and the Americas, demonstrating that the test can be administered meaningfully in vastly different cultural contexts -- though interpretation of scores across cultures remains a nuanced matter.

For those interested in testing their pattern recognition abilities, trying a practice IQ test offers exposure to matrix-style reasoning questions similar to those found on culture-fair assessments.

The Nature and Sources of IQ Test Bias

To understand what culture-fair tests aim to solve, it is essential to examine the specific mechanisms through which cultural bias enters intelligence testing.

Types of Test Bias

Psychometricians distinguish among several forms of bias:

Type of Bias	Definition	Example
Construct bias	The test measures different psychological constructs in different cultural groups	A "reasoning" test that actually measures English vocabulary for non-native speakers
Method bias	Differences in test administration, format, or response style affect groups unequally	Timed tests disadvantaging cultures that value deliberation over speed
Item bias (DIF)	Specific items function differently across groups after controlling for overall ability	A question about baseball rules on a test given in countries where cricket is played

Sources of Bias in Traditional IQ Tests

Bias enters intelligence testing through multiple channels:

Language dependence -- verbal instructions, reading passages, and vocabulary questions that require proficiency in the test language
Cultural content -- items referencing holidays, sports, customs, historical figures, or social norms specific to the dominant culture
Test-taking familiarity -- groups differ in their exposure to multiple-choice formats, timed tests, and the concept of standardized testing itself
Socioeconomic confounds -- access to education, nutrition, healthcare, and cognitively stimulating environments varies systematically across groups
Examiner effects -- the race, gender, or demeanor of the test administrator can influence performance through stereotype threat

"It is not that the tests are 'biased' in any simple sense; the problem is that they reflect the outcomes of social inequalities that already exist."
-- Claude Steele, Stanford University, researcher on stereotype threat

A landmark study by Steele and Aronson (1995) demonstrated that simply asking African American students to indicate their race before taking a standardized test significantly reduced their scores -- a phenomenon now known as stereotype threat. This effect operates independently of test content, showing that bias extends beyond the items themselves.

The Scale of the Problem

Research has documented score gaps on traditional IQ tests across multiple dimensions:

Factor	Typical Score Gap on Traditional Tests	Gap on Culture-Fair Tests
Socioeconomic status (high vs. low)	12-18 points	6-10 points
Native vs. non-native language speakers	10-15 points	3-7 points
Urban vs. rural populations	5-10 points	2-5 points
Western vs. non-Western cultural background	10-20 points	5-12 points

These data suggest that culture-fair tests reduce but do not eliminate group differences, a finding with important implications.

Do Culture-Fair Tests Truly Eliminate Cultural Bias?

The short answer is: they reduce it, but they do not eliminate it. The longer answer involves understanding why complete elimination may be conceptually impossible.

Evidence for Bias Reduction

Multiple research programs have demonstrated that culture-fair tests narrow score gaps:

Raven's Matrices shows smaller group differences than the Wechsler scales across racial and ethnic groups (Jensen, 1980; Rushton & Jensen, 2005)
Cattell's CFIT produces more similar score distributions across socioeconomic groups than verbally loaded tests (Cattell & Cattell, 1960)
Naglieri Nonverbal Ability Test (NNAT) was specifically designed to minimize score differences and shows reduced gaps in U.S. school populations (Naglieri & Ford, 2003)

Evidence for Persistent Bias

However, several lines of research demonstrate that nonverbal tests are not culture-free:

Test-taking strategies vary culturally -- some cultures emphasize speed while others emphasize accuracy, and timed tests inherently favor the former
Abstract reasoning itself is culturally trained -- exposure to puzzles, pattern games, and formal schooling builds the exact skills that "culture-fair" tests measure
The Flynn Effect is not culture-neutral -- IQ gains over time vary across countries and subtests, suggesting cultural factors drive changes even on nonverbal measures
Familiarity effects persist -- individuals who have never encountered multiple-choice formats or matrix-style problems perform worse initially, regardless of underlying ability

"There is no such thing as a culture-free test. The term 'culture-fair' is itself an aspiration rather than a description."
-- Robert Sternberg, former president of the American Psychological Association

Culture-Reduced, Not Culture-Free

The contemporary consensus among researchers is that the term "culture-reduced" is more accurate than "culture-free" or even "culture-fair." These tests minimize certain well-documented sources of bias -- especially language dependence and knowledge-based content -- but cannot eliminate the deeper ways that culture shapes cognition.

Practical implications of this distinction include:

Use culture-fair tests as one component of a broader assessment battery, not as standalone measures
Interpret results with cultural context -- a score of 110 on Raven's Matrices means something different for someone with 16 years of formal education than for someone with 4
Combine quantitative scores with qualitative evaluation -- interviews, work samples, and behavioral observations provide essential context
Use local norms when available rather than relying exclusively on norms derived from different populations

For those curious about how different test formats might impact performance, exploring a timed IQ test can illustrate how time pressure interacts with cognitive processing styles.

Other Major Culture-Fair Instruments

Beyond Cattell's CFIT and Raven's Matrices, several other instruments have been developed to address cultural bias in intelligence testing.

Comparison of Culture-Fair Test Instruments

Test	Developer	Year	Format	Best For
Raven's Progressive Matrices	John C. Raven	1938	Visual matrices	Research, general screening
Culture Fair Intelligence Test (CFIT)	Raymond B. Cattell	1949	Series, classification, matrices	Cross-cultural comparison
Naglieri Nonverbal Ability Test (NNAT)	Jack Naglieri	1997	Progressive matrices	School-age children, gifted identification
Leiter International Performance Scale (Leiter-3)	Roid & Miller	1929 (original)	Nonverbal subtests	Deaf, ESL, minimally verbal individuals
Test of Nonverbal Intelligence (TONI-4)	Brown, Sherbenou, Johnsen	1982	Abstract figural problems	Quick nonverbal screening
Universal Nonverbal Intelligence Test (UNIT-2)	Bracken & McCallum	1998	Six nonverbal subtests	Multilingual and cross-cultural assessment

The Naglieri Nonverbal Ability Test (NNAT)

Developed by Jack Naglieri at George Mason University, the NNAT was explicitly designed to reduce score gaps across racial and ethnic groups in U.S. schools. A study by Naglieri and Ford (2003) found that when the NNAT was used instead of traditional verbal-loaded tests, the number of African American students identified as gifted increased significantly -- from roughly 3% to 6% of identified students, more closely matching their proportion of the school population.

The Leiter International Performance Scale

Originally developed in 1929 by Russell Graydon Leiter for use with children who were deaf or spoke languages other than English, the Leiter-3 is entirely nonverbal in both its test items and its administration. Even the instructions are given through pantomime and demonstration, making it one of the most fully culture-reduced instruments available.

"The ideal test of intelligence would require no language, no reading, no writing, and no culturally specific knowledge. We can approach this ideal, but we cannot fully achieve it."
-- Alan Kaufman, co-developer of the Kaufman Assessment Battery for Children

Cross-Cultural Research: What the Data Show

Decades of cross-cultural research have produced a complex picture of how culture-fair tests perform around the world.

The Flynn Effect Across Cultures

The Flynn Effect -- the observation that average IQ scores have risen substantially over the 20th century -- provides important evidence about the relationship between culture and test performance. If IQ tests measured only innate biological ability, scores would remain stable over time. Instead:

Country	Period Studied	IQ Gain per Decade	Primary Test Used
Netherlands	1952-1982	+7.0 points	Raven's Matrices
United Kingdom	1938-2008	+2.5 points	Raven's Matrices
United States	1932-1978	+3.0 points	Stanford-Binet / Wechsler
Kenya	1984-1998	+11.0 points	Raven's CPM
Brazil	1930s-2002	+3.5 points	Various
Denmark	1959-2004	+3.0 points, then decline	Military test (Borge Priens Prove)

The fact that the Flynn Effect is largest on culture-fair tests like Raven's Matrices -- the very tests designed to measure "pure" fluid intelligence -- is paradoxical. It strongly suggests that environmental factors (nutrition, education, cognitive stimulation, urbanization) influence even nonverbal, abstract reasoning abilities.

"The gains are embarrassingly large on the very tests that were thought to be the most culture-free."
-- James R. Flynn, political scientist, University of Otago

Cross-Cultural Score Comparisons

When culture-fair tests are administered across diverse populations, the results must be interpreted carefully:

Within-group variation is always far larger than between-group variation -- individual differences within any cultural group dwarf average differences between groups
Score differences across nations correlate strongly with HDI (Human Development Index), years of schooling, and GDP per capita, suggesting environmental rather than innate causes
When environmental conditions equalize (as in transracial adoption studies or studies of immigrants' children), score gaps tend to narrow substantially

Challenges in Designing and Interpreting Culture-Fair Tests

Creating an effective culture-fair instrument requires navigating several interconnected challenges.

The Item Development Problem

Test developers must select stimuli that are:

Equally novel to all test-takers -- but prior exposure to puzzles, video games, and abstract art varies enormously across cultures
Free from culturally specific symbols -- even geometric shapes can carry cultural meaning (circles symbolize harmony in some East Asian contexts, triangles can have religious associations)
Appropriately difficult -- difficulty gradients may not be consistent across populations if different groups find different item types easier or harder

The Norming Problem

A culture-fair test is only as fair as its normative sample. If norms are derived primarily from Western, educated, industrialized populations -- what researchers Henrich, Heine, and Norenzayan (2010) called WEIRD (Western, Educated, Industrialized, Rich, Democratic) societies -- then applying those norms to non-WEIRD populations produces misleading results.

The Construct Validity Problem

Perhaps the deepest challenge: different cultures may conceptualize intelligence differently. Research across cultures has found:

East African concepts of intelligence (like the Luo concept of rieko) emphasize social responsibility and practical wisdom alongside cognitive speed
Chinese conceptions historically include moral character and self-knowledge
Brazilian street children demonstrate sophisticated mathematical reasoning in commercial contexts while performing poorly on formal mathematics tests

"What counts as intelligent behavior varies across cultural contexts, and any test that ignores this will inevitably measure cultural conformity alongside cognitive ability."
-- Patricia Greenfield, UCLA, cross-cultural psychologist

These findings do not invalidate culture-fair testing, but they do demand humility about what any single instrument can measure.

Practical Applications of Culture-Fair Testing

Despite their limitations, culture-fair tests serve critical roles in several domains.

Educational Placement

Culture-fair assessments help identify gifted students from underrepresented backgrounds who might be overlooked by traditional verbally loaded tests. The Naglieri and Ford (2003) study demonstrated this concretely: switching from the WISC to the NNAT for gifted screening increased minority representation in gifted programs without lowering the overall quality of identification.

Clinical and Neuropsychological Assessment

When evaluating individuals who are:

Non-native speakers of the test language
Deaf or hard of hearing
Minimally verbal due to autism, aphasia, or other conditions
Refugees or immigrants with disrupted educational histories

Culture-fair instruments like the Leiter-3 and TONI-4 provide more valid estimates of cognitive ability than language-dependent tests.

International Research and Cross-Cultural Comparison

Large-scale international studies -- such as Richard Lynn and Tatu Vanhanen's cross-national IQ analyses (though methodologically controversial) and the more rigorous PISA (Programme for International Student Assessment) -- rely on culture-reduced instruments to make meaningful comparisons across educational systems and national populations.

Workplace Assessment

Employers increasingly use nonverbal reasoning tests for hiring and promotion in multinational organizations, where candidates come from diverse linguistic and cultural backgrounds. Tests like the Wonderlic Personnel Test and various matrix reasoning assessments serve this purpose, though legal and ethical considerations require careful validation for each specific use case.

Future Directions in Fair Testing

The field of culture-fair testing continues to evolve, driven by advances in technology, psychometrics, and cross-cultural psychology.

Dynamic Assessment

Rather than measuring what a person already knows or can do, dynamic assessment measures learning potential -- how quickly and effectively someone acquires new skills with guided instruction. Pioneered by Reuven Feuerstein and grounded in Lev Vygotsky's concept of the zone of proximal development, dynamic assessment may be inherently more culture-fair because it focuses on the process of learning rather than its products.

Computerized Adaptive Testing (CAT)

Modern adaptive testing algorithms adjust item difficulty in real time based on the test-taker's responses, providing:

More precise measurement with fewer items
Reduced floor and ceiling effects
Potential for culturally adaptive item selection -- drawing from item banks calibrated for different populations

Neuroscience-Based Assessment

Emerging research explores whether direct measures of brain function -- such as neural processing speed, EEG coherence, or fMRI-measured network efficiency -- could provide culture-free indices of cognitive capacity. While promising in theory, these approaches remain experimental and raise their own questions about equity (access to scanning technology) and validity.

Game-Based Assessment

Interactive, game-like testing environments may reduce test anxiety and cultural unfamiliarity with formal testing formats. Companies like Arctic Shores and Pymetrics have developed game-based cognitive assessments for workplace use, though their psychometric properties are still being established.

"The future of fair testing lies not in finding the one perfect culture-free test, but in using multiple methods that together provide a more complete and equitable picture of human cognitive potential."
-- Robert Sternberg, Handbook of Intelligence (2020)

As you explore cognitive testing, consider trying a timed IQ test to experience how different formats can impact performance and reflect diverse cognitive processes.

Conclusion: Balancing Fairness and Validity in IQ Testing

Culture-fair IQ tests represent one of psychology's most important efforts to separate cognitive ability from cultural advantage. Instruments like Raven's Progressive Matrices and Cattell's CFIT have demonstrably reduced -- though not eliminated -- the influence of language, education, and cultural content on test performance. The contemporary consensus is clear: these are culture-reduced instruments, not culture-free ones.

The practical takeaway is equally clear: no single test should serve as the sole basis for consequential decisions about education, employment, or clinical diagnosis. The most equitable approach combines culture-fair instruments with verbal assessments, behavioral observations, and contextual understanding to build a comprehensive cognitive profile.

If you are interested in measuring your cognitive abilities with a balanced approach, you can take our full IQ test or start with a practice test to build familiarity. For a quicker experience, try our quick IQ assessment or challenge yourself with a timed IQ test.

By embracing both the promise and the limitations of culture-fair testing, we move closer to intelligence assessment that respects human diversity while maintaining scientific rigor.

References

Cattell, R. B. (1963). Theory of fluid and crystallized intelligence: A critical experiment. Journal of Educational Psychology, 54(1), 1-22.
Flynn, J. R. (2007). What Is Intelligence? Beyond the Flynn Effect. Cambridge University Press.
Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2-3), 61-83.
Jensen, A. R. (1998). The g Factor: The Science of Mental Ability. Praeger.
Naglieri, J. A., & Ford, D. Y. (2003). Addressing underrepresentation of gifted minority children using the Naglieri Nonverbal Ability Test (NNAT). Gifted Child Quarterly, 47(2), 155-160.
Raven, J., Raven, J. C., & Court, J. H. (2000). Manual for Raven's Progressive Matrices and Vocabulary Scales. Oxford Psychologists Press.
Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology, 69(5), 797-811.
Sternberg, R. J. (2020). The nature of intelligence and its development in childhood. In R. J. Sternberg (Ed.), Cambridge Handbook of Intelligence (2nd ed.). Cambridge University Press.
Greenfield, P. M. (1997). You can't take it with you: Why ability assessments don't cross cultures. American Psychologist, 52(10), 1115-1124.
Feuerstein, R., Feuerstein, R. S., & Falik, L. H. (2010). Beyond Smarter: Mediated Learning and the Brain's Capacity for Change. Teachers College Press.