The mind organizes the world into concepts — mental groupings of similar objects, events, and ideas ("furniture," "fairness"). For each concept we store a prototype: the best, most typical example of the category. Say "bird" and most Americans picture something robin-like, not a penguin; items closer to the prototype are recognized faster, while atypical members get judged slowly or miscategorized.
To solve problems, you have two basic tools. An algorithm is a step-by-step procedure that guarantees a solution — trying every possible combination on a lock will always work, eventually. A heuristic is a mental shortcut — fast, efficient, but with no guarantee (searching for keys "where I usually leave them"). The trade-off to memorize: algorithms are accurate but slow; heuristics are fast but error-prone. Sometimes neither runs to completion: insight is the sudden, all-at-once realization of a solution — the "Aha!" moment with no step-by-step path.
| Obstacle | Definition | Signature example |
|---|---|---|
| Mental set (a form of fixation) | Approaching a problem with the strategy that worked before, blocking a better one | Writing equations for a problem simple logic would solve |
| Functional fixedness | Seeing an object only in its usual function | Not using a heavy book as a doorstop |
| Confirmation bias | Seeking and remembering evidence that supports an existing belief, ignoring the rest | Noticing only a coworker's failures once you decide she's careless |
| Framing effect | Identical facts, different wording, different decision | "90% survival rate" chosen over "10% mortality rate" |
Two heuristics documented by Tversky and Kahneman deserve their own entries. The representativeness heuristic judges likelihood by resemblance to a prototype while ignoring base rates: a quiet, tidy detail-lover "must be" a librarian, even though salespeople vastly outnumber librarians. The availability heuristic judges likelihood by how easily examples come to mind: after vivid crash coverage, flying feels dangerous, though it is statistically far safer than driving. Keep them straight: representativeness = judging by resemblance; availability = judging by ease of recall.
Language is a system of symbols combined by rules. Learn the ladder from smallest to largest:
How do children acquire all this? B.F. Skinner proposed learning through imitation and reinforcement. Noam Chomsky countered with the decisive observation that children produce forms nobody models or rewards — a toddler says "goed" though every adult says "went." The child has extracted the -ed rule and is overapplying it, which imitation cannot explain. Chomsky argued humans are born with an innate language acquisition device primed to absorb the grammar of any surrounding language. The error "goed" is actually evidence of rule-learning, not regression.
Acquisition also appears to have a critical (sensitive) period: a window in early childhood when the brain is primed for language. Children exposed early reach native fluency; people who first tackle a language after that window — including adult second-language learners — typically build large vocabularies but rarely achieve fully native grammar and accent.
| Theory | Claim | Key evidence / example |
|---|---|---|
| Spearman's g | One general factor underlies all mental abilities | Scores on different mental tests correlate positively (found via factor analysis) |
| Gardner's multiple intelligences | Several relatively independent intelligences (linguistic, logical-mathematical, spatial, musical, bodily-kinesthetic, interpersonal, intrapersonal, naturalist) | Savant syndrome — one spectacular ability alongside low general functioning — is hard to square with a single g |
| Sternberg's triarchic theory | Three intelligences: analytical (school smarts), creative (novel ideas), practical (street smarts) | The colleague who aces every certification exam but can't manage a real project is strong analytically, weak practically |
Alfred Binet built the first practical test (Paris, 1905) around mental age — the age level a child's performance matches. Lewis Terman adapted it for American use as the Stanford-Binet, which reported an intelligence quotient: IQ = (mental age ÷ chronological age) × 100. An 8-year-old performing like a 10-year-old scores (10 ÷ 8) × 100 = 125. The formula breaks down for adults, so David Wechsler's tests — the WAIS for adults, the most widely used today — use a deviation IQ: your standing relative to your own age group, on a scale with a mean of 100 and standard deviation of 15.
A sound test needs three properties:
The relationship to memorize: reliability is consistency; validity is accuracy. A bathroom scale that always reads 10 pounds heavy is perfectly reliable and completely invalid. A test can be reliable without being valid — but never valid without being reliable.
Twin and adoption studies show intelligence has substantial heritability — a sizable share of the variation among individuals within a group is linked to genes. The critical reasoning move: heritability within a group says nothing about why two groups raised in different environments differ on average. Unequal schooling, nutrition, health care, and testing conditions all shift group averages without any genetic difference. Getting this within-group vs. between-group distinction right is the single most important safeguard in the debate over group differences in test scores.
1. E — A morpheme is the smallest unit that carries meaning (cat, plural -s, prefix un-). A) a phoneme is the smallest unit of sound — the number-one confusion; sound ≠ meaning. B) syllables are pronunciation chunks that often ignore meaning boundaries (unbreakable has four syllables but three morphemes). C) a grapheme is a written symbol, not a meaning unit. D) grammar is the entire rule system, not a minimal unit within it. Fix: phoneme = phone = sound; morpheme = "more meaning."
2. D — The most typical, best example of a category is its prototype, and typical members are recognized fastest. A) a heuristic is a problem-solving shortcut, not a category representation. B) an algorithm is a guaranteed solution procedure — nothing to do with categorization. C) the concept is the whole category "vehicle"; the car-like best example inside it is the prototype — the classic near-miss. E) a schema is a broad organizing framework of expectations, wider than the single best-example image. Fix: category = concept; the best example of that category = prototype.
3. B — Vivid, heavily covered crashes come to mind easily, inflating the felt probability of flying accidents — judging risk by ease of recall is the availability heuristic. A) representativeness is judging by resemblance to a prototype; no stereotype-matching occurs here. C) functional fixedness is about objects' uses, not risk estimates. D) an algorithm is a guaranteed procedure — Victor is doing the opposite of systematic calculation. E) confirmation bias would require an existing belief he selectively confirms; his fear was created by fresh vivid coverage. Fix: if a judgment tracks how easily examples pop to mind (not actual frequency), it's availability.
4. A — Sofia matches Tom to her librarian prototype and ignores the base rate (far more salespeople exist) — the signature of the representativeness heuristic. B) availability would require vivid examples coming easily to mind; the driver here is stereotype fit. C) mental set is repeating an old problem-solving strategy, not judging a person. D) insight is a sudden solution, not a categorization error. E) framing involves the same fact worded two ways; nothing was reworded. Fix: judging by resemblance to a stereotype while ignoring how common each category is = representativeness.
5. A — Andre is locked into the book's usual function (reading material) and misses its alternative use as a doorstop — functional fixedness. B) mental set is being stuck on a strategy that worked before; the block here is the object's function — the key contrast. C) confirmation bias concerns filtering evidence for a belief. D) framing concerns wording of options. E) insight is the escape from fixation (the sudden "use it as a wedge!"), not the obstacle itself. Fix: stuck on an object's usual use = functional fixedness; stuck on a former strategy = mental set.
6. C — Identical statistics worded as survival vs. mortality produce different choices — wording alone changed the decision, the framing effect. A) availability concerns ease of recalling examples, not wording of options. B) confirmation bias needs a prior belief being confirmed. D) representativeness is prototype-matching of category members. E) functional fixedness concerns objects' uses. Fix: same facts + different wording + different choices = framing.
7. B — "Goed" is never modeled or reinforced, so the child must have extracted the -ed rule herself — evidence for an innate rule-learning capacity (Chomsky's language acquisition device). A) is Skinner's position, which the unmodeled error directly undercuts — you cannot imitate a form nobody says. C) rule extraction happens years before formal schooling. D) memorization of separate forms would produce "went," not the invented "goed." E) phoneme inventory size is irrelevant to grammatical rule learning. Fix: unmodeled errors like "goed" = the child extracted a rule = evidence for innate acquisition, against pure imitation.
8. C — Early arrival brought full native grammar and accent; adult arrival brought vocabulary without native mastery — the dissociation predicted by a critical (sensitive) period for language. A) motivation can't explain the split: the father learned thousands of words, showing ample motivation. B) pure imitation predicts both learners would converge with enough exposure — twenty years should suffice. D) is exactly what the outcome refutes: vocabulary (learnable at any age) and grammar/accent (age-sensitive) dissociate. E) both have had decades of daily practice; hours alone don't close the gap. Fix: vocabulary comes at any age; native grammar and accent need the early window — that split = critical period.
9. E — Consistent scores on retest = high reliability; no relationship to the outcome it should predict = low (predictive) validity — the reliable-but-invalid test. A) is contradicted by the near-identical retest scores. B) reverses the terms: a test cannot be valid while unreliable, and here consistency is the strength. C) content validity concerns item coverage, and the reliability claim is wrong besides. D) consistent retest scores are direct evidence of reliability. Fix: reliability = consistency, validity = accuracy — a scale that's always 10 pounds off is reliable and invalid.
10. D — 85–115 is one standard deviation on each side of the mean (100 ± 15), which covers about 68% of a normal distribution. A) 34% is only one side of the band (mean to +1 SD). B) 50% is the share below the mean — a median confusion. C) 84% is the percentile below +1 SD, not the width of the band. E) 95% belongs to the ±2 SD band (70–130). Fix: ±1 SD ≈ 68%, ±2 SD ≈ 95% — check whether the question spans one or two SDs before answering.
11. D — Savant syndrome — one spectacular island of ability amid low general functioning — is Gardner's key evidence that abilities can be independent, which a single all-underlying g struggles to explain. A) positive correlations among tests are Spearman's own supporting evidence, not a challenge. B) predictive power of IQ is compatible with g. C) standardization is a testing procedure, silent on intelligence's structure. E) the normal distribution describes the spread of scores, not whether one factor produces them. Fix: evidence that abilities dissociate (savants, brain damage sparing one skill) challenges g; correlations among tests support it.
12. C — Heritability quantifies variation within a group; when two groups grow up in different environments, the between-group gap can be entirely environmental even if within-group heritability is high. A) is the precise error being tested — heritability does not transfer across the within/between line. B) is contradicted by large environmental effects (schooling, nutrition, enrichment) on IQ. D) overcorrects: the conclusion is flawed, but heritability of intelligence is real and substantial. E) reaches the right verdict for an unsupported reason — the flaw is the within/between leap, not a blanket claim that IQ tests measure nothing stable. Fix: within-group heritability says nothing about between-group differences when environments differ — the essential safeguard.
1. E — A morpheme is the smallest unit that carries meaning (cat, plural -s, prefix un-). A) a phoneme is the smallest unit of sound — the number-one confusion; sound ≠ meaning. B) syllables are pronunciation chunks that often ignore meaning boundaries (unbreakable has four syllables but three morphemes). C) a grapheme is a written symbol, not a meaning unit. D) grammar is the entire rule system, not a minimal unit within it. Fix: phoneme = phone = sound; morpheme = "more meaning."
2. D — The most typical, best example of a category is its prototype, and typical members are recognized fastest. A) a heuristic is a problem-solving shortcut, not a category representation. B) an algorithm is a guaranteed solution procedure — nothing to do with categorization. C) the concept is the whole category "vehicle"; the car-like best example inside it is the prototype — the classic near-miss. E) a schema is a broad organizing framework of expectations, wider than the single best-example image. Fix: category = concept; the best example of that category = prototype.
3. B — Vivid, heavily covered crashes come to mind easily, inflating the felt probability of flying accidents — judging risk by ease of recall is the availability heuristic. A) representativeness is judging by resemblance to a prototype; no stereotype-matching occurs here. C) functional fixedness is about objects' uses, not risk estimates. D) an algorithm is a guaranteed procedure — Victor is doing the opposite of systematic calculation. E) confirmation bias would require an existing belief he selectively confirms; his fear was created by fresh vivid coverage. Fix: if a judgment tracks how easily examples pop to mind (not actual frequency), it's availability.
4. A — Sofia matches Tom to her librarian prototype and ignores the base rate (far more salespeople exist) — the signature of the representativeness heuristic. B) availability would require vivid examples coming easily to mind; the driver here is stereotype fit. C) mental set is repeating an old problem-solving strategy, not judging a person. D) insight is a sudden solution, not a categorization error. E) framing involves the same fact worded two ways; nothing was reworded. Fix: judging by resemblance to a stereotype while ignoring how common each category is = representativeness.
5. A — Andre is locked into the book's usual function (reading material) and misses its alternative use as a doorstop — functional fixedness. B) mental set is being stuck on a strategy that worked before; the block here is the object's function — the key contrast. C) confirmation bias concerns filtering evidence for a belief. D) framing concerns wording of options. E) insight is the escape from fixation (the sudden "use it as a wedge!"), not the obstacle itself. Fix: stuck on an object's usual use = functional fixedness; stuck on a former strategy = mental set.
6. C — Identical statistics worded as survival vs. mortality produce different choices — wording alone changed the decision, the framing effect. A) availability concerns ease of recalling examples, not wording of options. B) confirmation bias needs a prior belief being confirmed. D) representativeness is prototype-matching of category members. E) functional fixedness concerns objects' uses. Fix: same facts + different wording + different choices = framing.
7. B — "Goed" is never modeled or reinforced, so the child must have extracted the -ed rule herself — evidence for an innate rule-learning capacity (Chomsky's language acquisition device). A) is Skinner's position, which the unmodeled error directly undercuts — you cannot imitate a form nobody says. C) rule extraction happens years before formal schooling. D) memorization of separate forms would produce "went," not the invented "goed." E) phoneme inventory size is irrelevant to grammatical rule learning. Fix: unmodeled errors like "goed" = the child extracted a rule = evidence for innate acquisition, against pure imitation.
8. C — Early arrival brought full native grammar and accent; adult arrival brought vocabulary without native mastery — the dissociation predicted by a critical (sensitive) period for language. A) motivation can't explain the split: the father learned thousands of words, showing ample motivation. B) pure imitation predicts both learners would converge with enough exposure — twenty years should suffice. D) is exactly what the outcome refutes: vocabulary (learnable at any age) and grammar/accent (age-sensitive) dissociate. E) both have had decades of daily practice; hours alone don't close the gap. Fix: vocabulary comes at any age; native grammar and accent need the early window — that split = critical period.
9. E — Consistent scores on retest = high reliability; no relationship to the outcome it should predict = low (predictive) validity — the reliable-but-invalid test. A) is contradicted by the near-identical retest scores. B) reverses the terms: a test cannot be valid while unreliable, and here consistency is the strength. C) content validity concerns item coverage, and the reliability claim is wrong besides. D) consistent retest scores are direct evidence of reliability. Fix: reliability = consistency, validity = accuracy — a scale that's always 10 pounds off is reliable and invalid.
10. D — 85–115 is one standard deviation on each side of the mean (100 ± 15), which covers about 68% of a normal distribution. A) 34% is only one side of the band (mean to +1 SD). B) 50% is the share below the mean — a median confusion. C) 84% is the percentile below +1 SD, not the width of the band. E) 95% belongs to the ±2 SD band (70–130). Fix: ±1 SD ≈ 68%, ±2 SD ≈ 95% — check whether the question spans one or two SDs before answering.
11. D — Savant syndrome — one spectacular island of ability amid low general functioning — is Gardner's key evidence that abilities can be independent, which a single all-underlying g struggles to explain. A) positive correlations among tests are Spearman's own supporting evidence, not a challenge. B) predictive power of IQ is compatible with g. C) standardization is a testing procedure, silent on intelligence's structure. E) the normal distribution describes the spread of scores, not whether one factor produces them. Fix: evidence that abilities dissociate (savants, brain damage sparing one skill) challenges g; correlations among tests support it.
12. C — Heritability quantifies variation within a group; when two groups grow up in different environments, the between-group gap can be entirely environmental even if within-group heritability is high. A) is the precise error being tested — heritability does not transfer across the within/between line. B) is contradicted by large environmental effects (schooling, nutrition, enrichment) on IQ. D) overcorrects: the conclusion is flawed, but heritability of intelligence is real and substantial. E) reaches the right verdict for an unsupported reason — the flaw is the within/between leap, not a blanket claim that IQ tests measure nothing stable. Fix: within-group heritability says nothing about between-group differences when environments differ — the essential safeguard.