Scroll to turn the page
Psychology · AP Psychology
perception, thinking, memory and intelligence
Cognition is everything the mind does with information: taking it in and making sense of it, thinking and deciding with it, storing it and getting it back, and the abilities intelligence tests try to measure.
The figure follows information through memory as the multi-store model pictures it. In each loop 12 stimuli arrive (an illustration, not data). Most fade from sensory memory within moments because no one paid attention to them. The 6 that are attended (red) enter short-term memory and are kept alive by rehearsal; 3 of them are encoded into long-term memory, from which retrieval can bring them back.
This chapter covers, in order:
The AP exam gives this unit 15–25% of its weight. Most questions describe a situation (a student studying, a witness, a test) and ask which concept it shows, or ask you to read a study’s design and data.
bottom-up and top-down processing; Gestalt grouping
Sensation is the senses detecting energy; perception is the brain organizing and interpreting it. The same input can be perceived in different ways.
Gestalt psychology (Wertheimer, Köhler, Koffka) held that the whole is different from the sum of its parts: the mind groups features into wholes by rules of its own. The figure takes 36 identical discs and changes one thing at a time:
Nothing about the discs changes from one grouping to the next: the grouping is done by the perceiver.
selective attention, the cocktail party effect, and what we fail to see
At any moment the senses take in far more than the mind can process. Selective attention focuses awareness on a small part of it, and the rest goes largely unnoticed.
In the figure, attention (the blue spotlight) follows the ball as it passes among the white team, 9 passes each loop. The gorilla (red) crosses the floor in plain view, outside the spotlight.
The lesson for the exam: we perceive what we attend to, not everything in front of us. Eyewitnesses, drivers using phones, and anyone absorbed in a task are open to it.
binocular and monocular cues; perceptual constancy; apparent movement
The image on each retina is flat, yet we see a world in three dimensions. The brain recovers depth from cues; infants show depth perception by the time they crawl (Gibson and Walk’s visual cliff, 1960).
Binocular cues need both eyes:
Monocular cues work with one eye, and are what painters use on a flat canvas (the picture in the figure):
Perceptual constancy keeps objects stable while their images change: size constancy (a person walking away does not seem to shrink), shape constancy (a door opening still looks rectangular), and lightness (color) constancy (white paper looks white in shadow). Apparent movement is motion seen where there is none, as in a film’s rapid series of still frames.
how thinking organizes the world
Thinking runs on concepts: mental groupings of similar objects, events, ideas or people. ‘Bird’ lets you know a thing you have never met will lay eggs and have feathers.
In the figure (an illustration, not measured distances) members of the concept sit at distances from the prototype by how typical they are. Each probe (blue) is judged by its likeness to the robin: the sparrow lands near and is judged fast, the penguin near the edge and slowly, and the bat outside.
Schemas are larger frameworks of concepts: what a restaurant visit or a classroom involves. Jean Piaget described two ways they change:
algorithms and heuristics; mental set and functional fixedness; creativity
Two strategies for a problem: follow a procedure that is sure to work, or take a shortcut that usually does.
In the figure the answer is in one of 16 boxes. The algorithm opens them in order and finds it on check 12; the heuristic tries the places experience points to (blue) and finds it on check 2. Had the answer been somewhere unlikely, the heuristic would have been the slow one.
Obstacles to solving problems:
Executive functions, largely in the frontal lobes, are the cognitive skills that manage problem solving: planning, organizing, holding goals in mind, shifting strategies and inhibiting impulses. They also make critical thinking possible: weighing evidence and questioning assumptions before reaching a conclusion.
Creativity is the ability to produce ideas that are both novel and valuable. It draws on divergent thinking, expanding the number of possible solutions (how many uses for a brick?), as well as convergent thinking, narrowing the possibilities to the single best one (a multiple-choice item).
heuristics, framing, priming and fallacies
Amos Tversky and Daniel Kahneman showed that quick judgments rely on heuristics that produce predictable errors.
The figure flips a fair coin. Before any flip, five heads in a row has probability ½5 = 3.1%. But once four have landed, the fifth is still 50% heads: the coin has no memory. The bars for the next flip never move, whatever the run.
Related errors you may meet: confirmation bias (seeking evidence that supports what you already believe) and overconfidence (being more confident than correct).
encoding, storage and retrieval; the multi-store model
Memory is learning that persists over time. It takes three processes: encoding (getting information in), storage (keeping it) and retrieval (getting it back out).
The multi-store model (Atkinson and Shiffrin, 1968) describes memory as three stages:
| store | holds | how long |
|---|---|---|
| sensory memory | a large, brief record of all sensory input | iconic (sight): a few tenths of a second; echoic (sound): 3–4 seconds |
| short-term memory | about 7 ± 2 items | under about 20 seconds without rehearsal |
| long-term memory | essentially unlimited | up to a lifetime |
Later researchers refined the model: short-term memory is an active working memory (next page), and some information reaches long-term memory automatically, without rehearsal. The levels of processing model (Craik and Lockhart, 1972) explains retention by how deeply information is processed rather than by which store it is in (page 2.4).
the working memory model; explicit and implicit memory; memory in the brain
Alan Baddeley and Graham Hitch (1974) recast short-term memory as working memory: a workspace that holds and manipulates information while you use it.
Because the loop and the sketchpad are separate, a verbal task and a visual task interfere with each other less than two verbal tasks do.
Long-term memory comes in two broad kinds:
| kind | what it is | example |
|---|---|---|
| explicit (declarative) | facts and experiences you consciously know and can declare | |
| · episodic | personal experiences, tied to a time and place | your first day of high school |
| · semantic | general facts and knowledge | Paris is the capital of France |
| implicit (nondeclarative) | retained without conscious recollection | |
| · procedural | skills and habits | riding a bike, typing |
Prospective memory is remembering to do something in the future (take medicine at 8 p.m.).
In the brain. The hippocampus is key to forming new explicit memories, which are later consolidated for storage in the cortex; the cerebellum and basal ganglia support implicit memories such as conditioned reflexes and skills; the frontal lobes support working memory. At the level of synapses, long-term potentiation (LTP) is the lasting increase in a synapse’s strength after brief, rapid stimulation, and is believed to be a neural basis for learning and memory.
automatic and effortful processing; levels of processing; chunking and mnemonics
Some information is encoded without trying; much takes effort, and how you encode it decides how well you will remember it.
Levels of processing (Craik and Lockhart, 1972): shallow processing encodes the surface, a word’s letters (structural) or its sound (phonemic); deep processing encodes its meaning (semantic). Deeper processing gives better retention. Relating material to yourself, the self-reference effect, is especially deep.
Ways to encode more:
primacy and recency; distributed practice
Learn a list and try to recall it at once: you remember the first and the last items best. This is the serial position effect.
The figure plots made-up data in the classic pattern: the percentage of people recalling each word of a 15-word list. Recalled at once (red), the curve is U-shaped: 78% for the first word, 35% in the middle, 90% for the last. After a delay filled with counting backward (blue), short-term memory has been cleared: the recency effect disappears and the primacy effect remains.
That pattern is evidence for separate short- and long-term stores. In a ‘which effect?’ question, ask which end of the list is favored and when recall happened.
The spacing effect: we retain information better when study is distributed over time (distributed practice) than when it is crammed into one session (massed practice), even for the same total time. Ebbinghaus noticed it in the 1880s. Spaced study with self-testing beats a night of rereading.
sensory, short-term and long-term memory; consolidation; amnesia
Each store has its own capacity and duration.
The figure shows the shape Peterson and Peterson (1959) found when they gave people three consonants and then had them count backward by threes so they could not rehearse (values illustrative): recall falls steeply within seconds, from 100% at once to about 11% after 18 seconds.
Maintenance rehearsal (simple repetition) holds information in short-term memory; elaborative rehearsal (linking it to what you know) helps move it to long-term storage.
Memory consolidation is the neural storage of a long-term memory: over time, and especially during sleep, memories formed with the help of the hippocampus are stabilized in the cortex. Autobiographical memory is memory for the events of one’s own life, a mix of episodic and semantic memory. A rare few people with highly superior autobiographical memory (hyperthymesia) can recall the events of almost every day of their lives.
When storage fails. Injury, disease and development can all impair storage. Amnesia is memory loss, usually from brain injury or disease:
recall, recognition and relearning; retrieval cues; context and state
A memory can be stored and still not come back when you need it. Three measures of retrieval show different amounts:
Retrieval cues are the anchors of memories: the more cues tied to a memory at encoding, the better. Priming activates associations that make related memories easier to reach. The encoding specificity principle says cues present at encoding are the best cues for retrieval, which explains the next three:
| learned | tested on land | tested underwater |
|---|---|---|
| on land | 13 | 9 |
| underwater | 8 | 12 |
The testing effect (retrieval practice): enhanced memory after retrieving information rather than simply rereading it. Metacognition, thinking about your own thinking, helps you judge what you actually know.
the forgetting curve; encoding failure, decay and retrieval failure
Hermann Ebbinghaus (1885) memorized lists of nonsense syllables (such as JIH and BAZ), then relearned them after delays. He measured what remained as savings: how much less time relearning took than the first learning.
| time since learning | 20 min | 1 hr | 9 hr | 1 day |
|---|---|---|---|---|
| savings | 58.2% | 44.2% | 35.8% | 33.7% |
| time since learning | 2 days | 6 days | 31 days |
|---|---|---|---|
| savings | 27.8% | 25.4% | 21.1% |
savings = (time to learn − time to relearn) ÷ time to learn
The forgetting curve: forgetting is rapid at first, then levels off. In his figures, savings fell to 58.2% within 20 min and to 33.7% within 1 day, but were still 21.1% after 31 days. The figure plots his points on a log scale of time.
Why we forget:
Overlearning (practicing beyond the first perfect recitation) increases retention, and spaced relearning flattens the curve.
proactive and retroactive interference; motivated forgetting
Memories compete. Learning one thing can disrupt recall of another, in either direction in time.
In the figure, list A was learned first and list B second. Tested on B, words from A push forward and some B words are lost (proactive); tested on A, B reaches back (retroactive). A memory trick: name the interference by the direction the disruption travels from the interfering learning.
Sigmund Freud proposed repression, the pushing of anxiety-arousing memories out of awareness; many researchers doubt that it happens as he described, since traumatic events are usually remembered all too well.
the misinformation effect, source amnesia and false memories
Memory is not a recording. Each time we retrieve a memory we rebuild it, and what we learned since can be woven in: reconsolidation and constructive memory.
The misinformation effect (Elizabeth Loftus): misleading information after an event distorts the memory of it. Loftus and Palmer (1974) showed students films of car crashes. Asked how fast the cars were going when they ‘smashed into’ each other, they estimated about 41 mph on average; asked about cars that ‘contacted’ each other, about 32 mph. In a second experiment, a week after the questions, all were asked whether they had seen broken glass. There was none in the film:
| question a week earlier | said yes to broken glass |
|---|---|
| ‘…when they smashed each other?’ | 16 of 50 (32%) |
| ‘…when they hit each other?’ | 7 of 50 (14%) |
| no question about speed | 6 of 50 (12%) |
One verb, a week earlier, more than doubled the false ‘yes’ answers: 32% against 14% and 12%.
For the exam: eyewitness testimony, leading questions in interviews, and ‘recovered’ memories of long-forgotten trauma are all open to these effects. Confidence in a memory is not proof that it is accurate.
g, multiple intelligences, the triarchic theory, fluid and crystallized
Intelligence is the ability to learn from experience, solve problems and use knowledge to adapt to new situations. Psychologists disagree about whether it is one ability or several.
Emotional intelligence (Salovey and Mayer) is the ability to perceive, understand, manage and use emotions.
mental age, IQ, standardization and the normal curve
Alfred Binet and Théodore Simon (1905) built the first practical test to find French schoolchildren who needed extra help. It estimated a child’s mental age: the level of performance typical of a given chronological age.
Lewis Terman at Stanford revised it for Americans (the Stanford-Binet, 1916) and used the intelligence quotient:
IQ = mental age ÷ chronological age × 100
A 8-year-old who performs like a typical 10-year-old has an IQ of 125. The ratio works poorly for adults, so modern tests such as David Wechsler’s WAIS (adults) and WISC (children) compare a person with others of the same age. The WAIS gives an overall score and subscores (such as verbal comprehension, perceptual reasoning, working memory and processing speed). Today IQ scores are often used to identify students for educational services, such as special education or gifted programs.
Standardization gives a test to a large, representative standardization sample under the same conditions, to set norms that scores are compared with. Scores are set so they fall in a normal curve with a mean of 100 and a standard deviation of 15:
| within | scores | share of people |
|---|---|---|
| 1 SD | 85–115 | 68% |
| 2 SD | 70–130 | 95% |
| 3 SD | 55–145 | 99.7% |
So about 2.5% score above 130 and as many below 70. An intellectual disability is diagnosed from a score of about 70 or below together with difficulty adapting to the demands of daily life; giftedness is usually set near 130.
The Flynn effect: average raw performance on intelligence tests rose through the 1900s, roughly 3 points per decade in many countries, so tests must be restandardized. The rise is linked to societal changes such as higher socioeconomic status, better health care and better nutrition. Achievement tests measure what you have learned (an AP exam); aptitude tests predict your ability to learn (the SAT was built as one).
what makes a good test; stereotype threat; mindset
A useful test must be both reliable (consistent) and valid (measuring what it claims to measure).
In the figure the bull’s-eye is the true value. Darts clustered on it: reliable and valid. Clustered off to one side: reliable but not valid (a scale that always reads five pounds heavy). Scattered: neither. A test can be reliable without being valid, but cannot be valid without being reliable.
Fairness and bias. A test is biased if it predicts less accurately for one group than another; items that depend on one culture’s experience can make it culturally biased. In the early 1900s, intelligence tests were misused in the United States to support the eugenics movement, including restrictions on immigration and forced sterilization. Scores have also been used to limit access to jobs, military ranks and schools.
Heredity and environment both matter: identical twins, even raised apart, have more similar scores than fraternal twins, and enriched or deprived environments raise or lower scores. Scores vary more within any group than between groups, and poverty, discrimination and unequal schooling can lower the scores of individuals and whole groups. Differences between groups do not show that they are genetic.