Cognitive Psychology Series Part 9: Intelligence & Individual Differences

1. History of Intelligence Testing

1.1 The Binet-Simon Scale (1905)

The first successful intelligence test was developed in 1905 by Alfred Binet and Theodore Simon in Paris. The French government commissioned Binet to create a tool to identify children who needed extra educational support — a practical, humanitarian goal. Binet developed a series of age-graded tasks and introduced the concept of mental age — a child performing at the level of an average 10-year-old was said to have a mental age of 10, regardless of their chronological age.

Critically, Binet himself warned against three misuses of his test that would all come to pass:

• Treating the score as measuring a single, fixed entity
• Using it to label children as permanently limited
• Using it to rank and sort people rather than to help them

1.2 The Wechsler Intelligence Scales

David Wechsler developed the most widely used intelligence tests in the world, beginning with the Wechsler-Bellevue in 1939. Unlike earlier tests that produced a single score, Wechsler's tests provide a profile of cognitive abilities across multiple domains.

1.3 Army Alpha & Beta Tests

2. Theories of Intelligence

2.1 Spearman's g Factor (1904)

Charles Spearman noticed that people who score well on one type of cognitive test tend to score well on all types. Using factor analysis (a statistical technique he helped pioneer), he proposed that all cognitive abilities are influenced by a single underlying factor — general intelligence (g) — plus specific factors (s) unique to each test.

2.2 Cattell's Fluid & Crystallized Intelligence

Raymond Cattell (1963) refined Spearman's g into two major components that we introduced in the context of aging in Part 8:

2.3 Sternberg's Triarchic Theory (1985)

Robert Sternberg argued that traditional IQ tests capture only a narrow slice of human intelligence. His triarchic theory proposes three interdependent aspects:

• Analytical intelligence: The ability to analyze, evaluate, compare, and judge — what traditional IQ tests measure well. Example: solving a logic puzzle or critically evaluating an argument.
• Creative intelligence: The ability to generate novel ideas, adapt to new situations, and see connections others miss. Example: inventing a new product, writing an original story, finding an unconventional solution.
• Practical intelligence: "Street smarts" — the ability to adapt to, shape, and select real-world environments. Example: knowing how to navigate office politics, negotiate a deal, or fix a car engine.

Sternberg's key criticism: a person could score average on an IQ test yet demonstrate remarkable practical or creative intelligence in real-world contexts. Traditional tests, he argued, systematically undervalue these forms of intelligence.

2.4 Gardner's Multiple Intelligences (1983)

Howard Gardner proposed the most radical departure from the g-factor tradition. He argued that intelligence is not a single ability but a set of eight relatively independent intelligences:

3. Emotional Intelligence

Emotional Intelligence (EI) — first formally proposed by Peter Salovey and John Mayer (1990) and popularized by Daniel Goleman (1995) — refers to the ability to perceive, understand, manage, and use emotions in oneself and others.

The Salovey-Mayer Four-Branch Model organizes EI into a hierarchy of abilities:

4. Cognitive Styles

Cognitive styles describe how people process information — their characteristic patterns of perceiving, thinking, remembering, and problem-solving. Unlike abilities (which vary in amount), styles are preferences that influence approach but not necessarily success.

5. Nature vs Nurture

Perhaps no question in intelligence research generates more heat than the nature-nurture debate. The answer, as with most false dichotomies in psychology, is "both — and they interact in complex ways."

5.1 Twin & Adoption Studies

The most powerful designs for disentangling genetic and environmental influences are twin studies (comparing identical vs fraternal twins) and adoption studies (comparing adopted children with biological and adoptive parents).

5.2 The Flynn Effect

# Visualizing IQ Distribution and the Flynn Effect
import numpy as np

class IQAnalyzer:
    """
    Analyzes IQ score distributions, correlations between measures,
    and simulates the Flynn Effect across generations.
    """

    def __init__(self, mean=100, sd=15):
        self.mean = mean
        self.sd = sd

    def generate_distribution(self, n=10000):
        """Generate a normal distribution of IQ scores."""
        scores = np.random.normal(self.mean, self.sd, n)
        return scores

    def classify_scores(self, scores):
        """Classify IQ scores into standard categories."""
        categories = {
            'Extremely Low (< 70)': np.sum(scores < 70),
            'Borderline (70-79)': np.sum((scores >= 70) & (scores < 80)),
            'Low Average (80-89)': np.sum((scores >= 80) & (scores < 90)),
            'Average (90-109)': np.sum((scores >= 90) & (scores < 110)),
            'High Average (110-119)': np.sum((scores >= 110) & (scores < 120)),
            'Superior (120-129)': np.sum((scores >= 120) & (scores < 130)),
            'Very Superior (130+)': np.sum(scores >= 130)
        }
        return categories

    def simulate_flynn_effect(self, start_year=1920, end_year=2020,
                               gain_per_decade=3.0):
        """Simulate the Flynn Effect: IQ gains over time."""
        decades = (end_year - start_year) / 10
        years = list(range(start_year, end_year + 1, 10))
        means = [self.mean + gain_per_decade * (y - end_year) / 10
                 for y in years]

        print("=== Flynn Effect Simulation ===")
        print(f"Gain rate: {gain_per_decade} IQ points per decade")
        print(f"\n{'Year':<8}{'Mean IQ (by today norms)':<28}{'Relative to 2020'}")
        print("-" * 50)
        for year, mean_iq in zip(years, means):
            diff = mean_iq - self.mean
            bar = "█" * max(0, int((mean_iq - 60) / 2))
            print(f"{year:<8}{mean_iq:<28.1f}{diff:+.1f}")

        total_gain = gain_per_decade * decades
        print(f"\nTotal gain ({start_year}-{end_year}): {total_gain:.0f} IQ points")
        return years, means

    def correlate_measures(self):
        """
        Simulate correlations between different cognitive measures
        to demonstrate the positive manifold (Spearman's g).
        """
        n = 1000
        # Generate a common g factor
        g = np.random.normal(0, 1, n)

        # Each test = g + specific factor + noise
        tests = {
            'Vocabulary': g * 0.7 + np.random.normal(0, 0.7, n),
            'Matrix Reasoning': g * 0.75 + np.random.normal(0, 0.65, n),
            'Digit Span': g * 0.55 + np.random.normal(0, 0.83, n),
            'Processing Speed': g * 0.5 + np.random.normal(0, 0.87, n),
            'Spatial Rotation': g * 0.65 + np.random.normal(0, 0.76, n)
        }

        print("\n=== Positive Manifold: Correlation Matrix ===")
        print("(All cognitive tests correlate positively with each other)")
        test_names = list(tests.keys())
        print(f"\n{'':20s}", end='')
        for name in test_names:
            print(f"{name[:8]:>10s}", end='')
        print()

        for i, name_i in enumerate(test_names):
            print(f"{name_i:20s}", end='')
            for j, name_j in enumerate(test_names):
                corr = np.corrcoef(tests[name_i], tests[name_j])[0, 1]
                print(f"{corr:10.2f}", end='')
            print()

        print("\nAll correlations are positive — this is the 'positive manifold'")
        print("that supports Spearman's g factor theory.")

# Run analysis
analyzer = IQAnalyzer()
scores = analyzer.generate_distribution(10000)

print("=== IQ Score Distribution (n=10,000) ===")
categories = analyzer.classify_scores(scores)
for cat, count in categories.items():
    pct = count / len(scores) * 100
    bar = "█" * int(pct / 2)
    print(f"  {cat:30s}: {count:5d} ({pct:5.1f}%) {bar}")

print()
analyzer.simulate_flynn_effect()
analyzer.correlate_measures()

6. Stereotype Threat

Stereotype threat, discovered by Claude Steele and Joshua Aronson (1995), is the anxiety that arises when a person is at risk of confirming a negative stereotype about their social group. This anxiety consumes working memory resources and impairs performance — creating a self-fulfilling prophecy.

How stereotype threat impairs performance:

• Working memory drain: Anxiety and self-monitoring consume limited working memory resources needed for the task
• Increased arousal: Physiological stress response impairs complex cognitive performance
• Prevention focus: Shifting from trying to succeed to trying not to fail (more cautious, slower, avoidant)
• Reduced sense of belonging: Feeling like an outsider in the domain

Effective interventions:

• Framing tests as non-diagnostic of ability
• Emphasizing that intelligence is malleable (growth mindset intervention)
• Values affirmation exercises before high-stakes tests
• Providing role models who counter the stereotype
• Teaching students about stereotype threat itself (metacognitive awareness)

7. Growth vs Fixed Mindset

Carol Dweck's research on implicit theories of intelligence has transformed education and organizational psychology. Her central finding: what people believe about intelligence profoundly affects their motivation, learning behavior, and ultimately their achievement.

8. Giftedness, Intellectual Disability & Extremes of Intelligence

The extremes of the intelligence distribution — both high and low — raise important questions about the nature of cognitive ability and how educational systems should respond to individual differences.

Giftedness and Gifted Education

Gifted individuals are typically identified as scoring in the top 2-3% on standardized intelligence tests (IQ 130+), though modern definitions increasingly include exceptional talent in specific domains regardless of overall IQ score.

Intellectual Disability

Intellectual disability (formerly called "mental retardation") is defined by three criteria that must all be present: (1) significant limitations in intellectual functioning (IQ approximately 70 or below), (2) significant limitations in adaptive behavior, and (3) onset during the developmental period. Approximately 1-3% of the population meets these criteria.

Exercises & Self-Assessment

Conclusion & Next Steps

In this chapter, we have explored one of psychology's most fascinating and contentious topics — the nature and measurement of human intelligence. Here are the key takeaways:

• Intelligence testing has a complex history — from Binet's humanitarian goals to the misuse of Army tests for immigration restrictions. Context and cultural fairness matter enormously
• Spearman's g factor is robustly supported by the positive manifold, but theorists like Sternberg and Gardner argue that traditional IQ tests capture only a narrow slice of human cognitive ability
• Fluid intelligence peaks in the mid-20s and declines, while crystallized intelligence continues growing through the 60s — intelligence is not a single number that defines you
• Emotional intelligence predicts success in social and leadership domains beyond what IQ captures, though the popular press has exaggerated its importance
• Both genes and environment shape intelligence, with heritability increasing across development due to gene-environment correlation. The Flynn Effect demonstrates that environmental changes can produce massive IQ gains
• Stereotype threat shows that test performance is not just about ability — it is profoundly influenced by social context and identity
• Growth mindset — the belief that intelligence can be developed — promotes resilience, effort, and achievement, especially when translated into concrete behavioral changes