Psychology curriculum

Psychologists study thought, emotion and behavior, using observation, assessment and experiment. Calling the field scientific commits it to evidence other people can check. Common sense and introspection alone can mislead, because many mental operations are not open to conscious inspection.

Wundt's laboratory, formally recognized in 1879, aimed at an exact description of conscious experience; Titchener trained observers at length to report it. James looked inward too, but asked what the mind is for, treating the pursuit of goals as the mark of a mind. Trained observers could not settle disputes such as imageless thought, which helped behaviorism push introspection aside.

Popper argued that psychoanalysis could explain any behavior at all, so no observation could ever count against it, which made its reach a weakness rather than a strength. Later critics pressed a sharper point: the real threat to its scientific standing is not that it cannot be tested but that its claims have been tested and failed.

Watson held that private mental states cannot be objects of a public science, so psychology should study behavior; Skinner explained behavior through its environment and history of reinforcement. The gains survive in behavior therapy and animal learning research. The cost was ignoring how a person represents a situation, the main reason the program lost its dominance.

Humanistic psychology was called the third force because it held that neither behaviorism nor psychoanalysis covered the full range of human possibility; Maslow founded its association in 1961. Rogers proposed an actualizing tendency, a built-in push toward growth and autonomy that a poor social environment can thwart, and he studied his approach with ordinary empirical methods.

Behaviorism had banished talk of mental representation, but around 1956 the picture changed. Miller argued that short-term memory holds only about seven items, early computers launched artificial intelligence, and Chomsky explained language with mental rules rather than learned habits. Mental processes returned as models tested by experiment and by computer simulation.

Engel's 1977 model holds that biological, psychological and social factors all causally influence health and illness, so no single level tells the whole story. Its main criticism is vagueness: because it names no method, a clinician can stress whichever level they already favor. A structured formulation that walks through every level is the usual guard against that.

Clinical and counseling psychologists assess and treat mental, emotional and behavioral disorders. Industrial-organizational psychologists apply research to work, such as hiring and productivity; school psychologists tackle problems that affect learning; forensic psychologists help courts with the psychological side of a case; neuropsychologists study how brain damage changes behavior. Each subfield is defined by the question it asks.

Psychiatrists are physicians, so they can diagnose and treat mental illness with medication as well as therapy. Clinical psychologists usually hold a PhD or PsyD, are licensed by each state, and treat through assessment and psychotherapy; only some states let them prescribe. Both offer psychotherapy, which is why training and prescribing authority, not the talking, separate them.

You test a theory by deducing a specific prediction from it and checking whether the result falsifies or corroborates it. Popper's test of a scientific theory is risk: it must rule out some possible observations. A theory compatible with every possible result, or one rescued from each failure by an added ad hoc assumption, has stopped making testable claims.

In an observational study you watch or measure people without trying to change the outcome. Observing behavior in its natural setting captures complexity that simplified laboratory tasks miss, but you give up control over what varies. A case study of one person goes deep, yet the data alone cannot tell you whether a pattern is typical or unique to that individual.

Correlational data, gathered without any manipulation, describe how measures go together and can support prediction. They cannot settle direction: deviant friends may increase deviant behavior, or the reverse. They also face the third-variable problem, where an unmeasured common cause produces a spurious correlation or masks a real one.

In an experiment you manipulate the presumed cause, the independent variable, and measure its effect on the dependent variable. Random assignment makes groups equal on average, which is what removes systematic confounding. It guarantees nothing for one study, though: on any baseline measure tested at the .05 level, about one study in twenty will show a significant difference before treatment begins, which a pretest can catch.

A control group skips the treatment and gives you a baseline to compare against. A placebo looks and is given exactly like the real treatment, so both groups share the same expectations. In a single-blind study participants do not know which condition they are in; in a double-blind study the researchers do not know either, which closes off bias from both sides.

Random assignment protects a study's internal validity, but whether results generalize depends on who was sampled in the first place. Studies from the US find that WEIRD participants (Western, educated, industrialized, rich, democratic) make up roughly 70 to 90 percent of psychology samples, often students or online volunteers chosen for convenience, so a finding may not hold even for other Americans.

Demand characteristics are the cues that tell participants what a researcher expects. Orne noticed how helpful people were: many tried to play the good subject and give the result they thought the study needed. Laboratory studies therefore often hide the true purpose through blinding or distraction, a protection that is much harder to arrange in field studies.

Psychological constructs cannot be measured directly, so you study them through scores from a chosen procedure. That choice limits the conclusion: two procedures can give similar numbers and still capture different things, so whether they measure one concept has to be shown, not assumed. Defining a concept as nothing but its procedure is now widely seen as an extreme position.

Reliability is consistency: similar items should agree, and a score should not swing with random error. Validity is whether the score reflects the construct it claims to measure. Reliability usually comes first, since an inconsistent score cannot track a stable trait, but it is not enough: a measure can be consistent and still miss its target.

The mean adds the values and divides by their number, the median is the middle value once they are ranked, and the mode is the most frequent one. In a perfectly symmetric bell curve all three coincide. A few extreme values pull the mean toward them, so a mean above the median signals right skew, and the median then represents the data better. Spread is a separate question, measured by the range, interquartile range or standard deviation.

Pearson's r runs from -1 to +1 and tells you how closely the points fit a straight line, not how steep that line is. It assumes a linear pattern and is overly sensitive to outliers: in Anscombe's quartet, four data sets share the same r, yet one is curved and one is driven by a single extreme point. Plot the scatterplot before you trust the number, and remember that an association alone does not show causation.

A p-value is the probability of data at least as extreme as yours if every assumption of the test, null hypothesis included, were true. It is not the chance the null is true, nor the chance a significant result is a false positive. It does not measure importance, since a large study can make a trivial effect significant, and a nonsignificant result does not prove the effect is small.

Effect size says how big a difference is; power is the chance, before the study, that it will detect an effect that is really there. A typical psychology effect needs about 99 people per group for 80% power in a two-group design, a size studies rarely reach. Small studies give unstable p-values, and because only large observed effects reach significance in them, published effects tend to be inflated.

When researchers repeated 100 studies from three leading psychology journals, well under half came out significant again. Proposed causes include publication bias, p-hacking (such as stopping data collection once significance appears) and HARKing, presenting an unexpected result as if it was predicted. Reforms include preregistration, a public, date-stamped statement of hypotheses, plus data sharing and multi-lab projects such as Many Labs.

Under the US Common Rule, consent must be sought in conditions that minimize coercion or undue influence, in language the person understands, with the information a reasonable person would want before deciding. Participation is voluntary and can be stopped at any time without penalty. A consent form may not include language that makes participants waive their legal rights.

US rules let a review board alter consent, including about a study's purpose, only when the research is minimal risk, could not practicably be done otherwise, and does not harm participants' rights and welfare. When appropriate, participants get the missing information afterward, which is the debriefing. One exemption instead requires people to agree in advance that they may be misled.

An institutional review board examines a study before it runs and can approve it, require changes, or reject it. To approve, it must find that risks are minimized through sound design and are reasonable in relation to the expected benefits and the importance of the knowledge. A board that often reviews vulnerable groups, such as children or prisoners, should consider including members with relevant expertise.

A Certificate of Confidentiality, used in much US research, bars researchers from disclosing identifiable, sensitive data to anyone outside the study or in legal proceedings. The protection has narrow exceptions: disclosure required by other laws, such as reporting communicable disease or child or elder abuse, release with the participant's consent, and disclosure needed for the participant's medical treatment.

Milgram ran his obedience studies in 1961-62, before review boards existed: in the baseline version 65% gave the maximum fake shock, though across about 30 variants most participants stopped short. Zimbardo's prison study was halted after six days of guard abuse. The Tuskegee syphilis study ran from 1932 to 1972 without informed consent and withheld treatment, and it drastically changed research practice.

US government principles require that animal procedures be justified by their relevance to health, knowledge or the good of society. Researchers should use the fewest animals that give valid results, consider alternatives such as computer models, and avoid or minimize pain, assuming that what hurts a person may hurt an animal. Exceptions are decided by a review committee, not by the investigators.

Dendrites branch out and receive messages, the cell body holds the nucleus and makes most of what the neuron needs, and a single axon carries the signal away to another neuron, a muscle or a gland. Many axons are wrapped in fatty myelin, which speeds signals, with gaps called nodes of Ranvier. The axon's branches end in small swellings at the synapse.

At rest the membrane lets potassium through far more easily than sodium, so potassium's concentration difference dominates the resting charge. The sodium-potassium pump keeps those gradients in place, spending one ATP to move three sodium ions out and two potassium ions in. When the pump falters, the gradients run down and the cell's excitability is disturbed, a mechanism studied in epilepsy.

When enough input pushes the membrane to threshold, the neuron fires an action potential: a brief spike of about 50 to 100 millivolts that depends on voltage-gated sodium and potassium channels and usually starts where the axon leaves the cell body. Potassium current brings the membrane back down. The spike is all-or-none, so input that falls short of threshold produces no spike at all.

When a signal reaches the end of an axon, vesicles release neurotransmitter into the synaptic cleft; it crosses and binds receptors on the next cell, like a key in a lock, changing that cell's activity. Then the signal is shut off. Transporters pull many transmitters back into the sending neuron, which is reuptake, while acetylcholine is broken down in the cleft by an enzyme, acetylcholinesterase.

Excitatory transmitters such as glutamate make a receiving cell more active; inhibitory ones such as GABA dampen it. A neuron fires only when its combined input is strong enough, and inhibition is an active part of that sum, not just missing excitation: dedicated GABA-releasing interneurons send signals of their own, producing inhibitory postsynaptic potentials that push the cell away from firing.

Glutamate is the main excitatory transmitter and GABA a key inhibitory one. Acetylcholine triggers muscle contraction, and its shortage is linked to Alzheimer's disease. Serotonin is involved in sleep and temperature, dopamine in mood, movement and reinforcement, with dopamine loss behind Parkinson's rigidity. Norepinephrine works both as a transmitter and as a hormone released under stress.

An agonist binds a receptor and produces the same action as the substance that normally binds there; an antagonist stops another substance's effect. Drugs can also act at other points in the cycle. Heroin and marijuana mimic natural transmitters at receptors, cocaine blocks the dopamine transporter so dopamine builds up in the synapse, and SSRIs slow the reabsorption of serotonin.

The central nervous system is the brain and spinal cord; the peripheral system is the nerves and ganglia outside them. Peripheral nerves carry sensory signals in and motor signals out. The motor side splits again: the somatic system drives skeletal muscles under voluntary control, while the autonomic system runs heart muscle, smooth muscle and glands without conscious effort.

Many organs receive fibers from both autonomic divisions, with one stimulating and the other inhibiting. In a crisis, parasympathetic influence withdraws and sympathetic outflow rises for fight or flight; parasympathetic, largely vagal, influence supports rest and digestion. Neither system ever fully switches off: even at the extremes the two stay in a constant tug-of-war.

The hindbrain includes the brainstem and the cerebellum and controls vital functions such as breathing and heart rate. Every sensory and motor fiber linking brain and spinal cord passes through the medulla; the pons carries conduction tracts and helps produce REM sleep's muscle relaxation. The cerebellum coordinates and refines learned movements, and the midbrain controls some reflexes and eye movements.

The hippocampus works as a memory indexer, sending memories to the cortex for long-term storage and retrieving them. The hypothalamus governs drives such as hunger, thirst and body temperature and helps trigger the stress response. The thalamus relays sensory signals, except smell, to the cortex, and the amygdala is involved in processing emotion.

The frontal lobes handle planning and reasoning and hold one idea in mind while you consider others; the motor cortex at their rear plans and executes voluntary movement. The parietal lobes take in touch, temperature, taste and movement from the body. The occipital lobes process vision, so damage there can cause blindness. The temporal lobes receive sound and help form and retrieve memories.

Split-brain patients had the corpus callosum cut to relieve severe epilepsy, which let researchers test each hemisphere on its own. Some functions do lean to one side; forming words relies mainly on the left. The popular idea of left-brained and right-brained people fails, though: a scan study of 1,011 people found lateralization varies network by network, not as a whole-brain type.

Experience changes the brain's wiring: learning alters connections between neurons, adult rats grow new hippocampal neurons, and canaries add forebrain neurons in the season they learn new songs. Whether adult humans do the same, and how much, is still intensely debated. Repair has limits too: fibers inside the brain and spinal cord lack the sheath that guides regrowth, so central pathways do not regenerate the way peripheral nerves can.

EEG tracks neural activity millisecond by millisecond, while fMRI shows which areas are active in fine spatial detail but has its timing blurred by the slow blood-oxygen response. PET follows radioactive tracers through the brain. Beware reverse inference: concluding that a mental process occurred because a region lit up fails when that region serves other functions too.

Your genotype is the specific variant you carry at a given place in the genome; your phenotype is what can be observed, from height and eye color to the presence of a disease. A phenotype is shaped by genotype and environment together, and there is usually no one-to-one match between them, because environmental effects can equal or outweigh genetic ones.

Heritability is the share of variation in a trait across a population that can be attributed to genetic variation. It describes a group, not a person, so a heritability of one half does not mean half of your trait comes from your genes. It also says nothing about how genes build a trait during development, and gene-environment interaction makes even the population split hard to pin down.

Twin studies compare identical twins, who share nearly all their DNA, with fraternal twins, who share about half, assuming both kinds share their environment equally; extra likeness in identical twins is credited to genes. Adoption studies split the two another way: resemblance to birth parents who did not raise the child points to genes, resemblance to adoptive parents to rearing. Each estimate is only as good as its assumptions.

A gene-environment interaction exists when the same environment affects different genotypes differently. A well-known example: in 2002 Caspi's team found that childhood maltreatment predicted later antisocial behavior in males carrying a low-activity version of the MAOA gene. Two later meta-analyses supported the finding, though candidate-gene results of this kind remain methodologically debated.

Epigenetic changes are chemical marks on DNA and its packaging proteins that change how strongly genes are switched on, without changing the DNA sequence. Some marks pass to daughter cells when cells divide. Studies in animals and people link early-life stress to such marks on the gene for the glucocorticoid receptor, which helps regulate the stress response, one route by which experience may leave a lasting trace.

Glands secrete hormones into the blood, which carries them everywhere, yet only cells with matching receptors respond. Their effects take minutes to weeks and spread widely, unlike the seconds-long, local action of nerve signals. The hypothalamus steers the anterior pituitary through releasing hormones, and the pituitary's hormones in turn direct the thyroid, adrenal glands and gonads; many hormones are held in check by negative feedback.

Clinicians distinguish several states between coma and full awareness. A patient with unresponsive wakefulness has sleep-wake cycles but shows no sign of awareness, while a minimally conscious patient seems to perform intentional actions. Brain scans have shown signs of following commands in some patients diagnosed as unresponsive, which is why bedside behavior can understate awareness.

Your biological clock runs on a roughly 24-hour cycle and shapes wakefulness, body temperature, metabolism and hormone release. Its master control, the suprachiasmatic nucleus in the hypothalamus, gets light information directly from the eyes and drives the pineal gland's melatonin as darkness falls. Light is the strongest cue, which is why shift work and jet lag put the internal clock out of step.

Sleep cycles between three stages of non-REM sleep and REM sleep roughly every 80 to 100 minutes, four to six times a night. Stage 3, deep slow-wave sleep, comes in longer stretches early in the night, while REM periods grow longer later on. Stage 2, a lighter sleep marked by brief bursts of brain activity, takes up the most time across the night's cycles.

REM sleep first arrives about 90 minutes after you fall asleep. Your eyes dart, brain waves resemble waking, breathing and heart rate rise, and arm and leg muscles are temporarily paralyzed so you do not act out your dreams. In REM sleep behavior disorder that paralysis fails and people enact their dreams, sometimes injuring themselves or a bed partner.

Sleep's biological purpose is still not settled, so several accounts coexist. Sleep may clear toxins that build up in the brain while you are awake, and memory consolidation most likely needs both non-REM and REM sleep. The synaptic homeostasis hypothesis adds that waking strengthens synapses and sleep scales them back; a rival sequential hypothesis argues that sleep's main memory job is choosing which traces to keep.

Too little sleep makes it harder to learn, focus and react, leaves people cranky or worried, and is linked to more infections and to diabetes and obesity. Whether people notice their own decline is less clear than often assumed: in one trial of 182 adults, a night without sleep only slightly reduced how well people judged their own test performance.

Freud treated dream interpretation as the royal road to the unconscious. Hobson countered that dreams arise from random, brainstem-driven activation during sleep and offer insight at best the way an inkblot test does. The continuity hypothesis, backed by analyses of dream content, holds that dreams follow the dreamer's waking concerns and memories. Weigh each by what systematic study of dream reports supports.

With repeated exposure the brain adapts, for example by making fewer neurotransmitters or fewer receptors in the reward circuit, so the original dose does less and a larger one is needed: that is tolerance. Stop the drug abruptly and the adaptation is left unopposed, producing withdrawal. With depressants, which slow brain activity, the rebound can include seizures.

Dependence is the body's adaptation to long-term use, revealed by withdrawal when use stops suddenly, and it can usually be managed by tapering the dose. Addiction is a separate process: compulsive seeking and use despite harm, driven by changes in brain circuits. Someone taking a prescribed opioid for weeks can become dependent without being addicted, while drug cues can trigger cravings years after withdrawal is over.

Benzodiazepines, barbiturates and the z-drug sleep aids slow brain activity, mostly by boosting signaling at receptors for GABA, the brain's main inhibitory messenger. Barbiturates carry a higher overdose risk than benzodiazepines. The added danger comes from combining sedating drugs: alcohol taken with sleep or anti-anxiety medicines or with opioids raises the risk of overdose, because their effects stack.

Stimulants raise alertness, attention and energy along with heart rate and blood pressure. Cocaine blocks the transporter that clears dopamine, amphetamine-type drugs amplify dopamine and norepinephrine signaling, and nicotine raises dopamine in reward circuits, a surge that reinforces taking the drug. Caffeine works differently: it keeps you alert by blocking adenosine, a chemical that builds sleepiness.

Opioid drugs attach to and activate opioid receptors in the brain, spinal cord and gut, blocking pain signals; the body makes its own opioids, the endorphins. They also act on reward regions, causing euphoria, and on brainstem areas that control breathing, so too much can stop breathing altogether. Combining opioids with alcohol or sedatives raises that risk, and naloxone can reverse an overdose if given quickly.

Psychedelics such as psilocybin and LSD act mainly on how the brain processes serotonin and can bring vivid visions and shifts in the sense of self. Dissociatives such as ketamine and PCP act mainly on glutamate and can make people feel detached from their body and surroundings. The experience depends on dose and potency, on mood and expectations (the set) and on the environment (the setting).

People differ widely in hypnotic suggestibility, measured with standard scales; researchers usually count the top 10 to 15 percent as highly hypnotizable. Suggestions can change how voluntary an action feels and alter perception. Mindfulness trains attention on the present. One large analysis found it beat no treatment for diagnosed anxiety and depression and matched established therapies; a review put negative effects at about 8 percent.

Sensation is the receiving end: the eye's photoreceptors turn light into electrical signals, and the optic nerve carries them away. Perception is what the brain then makes of those signals, turning them into the images you see. The split matters because perception can go beyond the input, as when a few wedge-shaped gaps make you see a triangle that is not drawn.

Each sense has its own converter from physical energy to neural signal. In the eye, photoreceptors turn light into electrical signals. In the ear, sound sets the cochlear fluid rippling; tiny projections on hair cells bend, opening channels so that charged particles rush in and create an electrical signal. Smell and taste begin when molecules stimulate receptors on specialized sensory cells.

The absolute threshold is the weakest stimulus that can be sensed. For human vision under ideal conditions, estimates run from a single photon absorbed by one rod to single photons in about a dozen rods. Much of that spread comes from decision bias, an observer's tendency to say yes or no regardless of the stimulus, which is why careful methods measure false alarms as well.

The just-noticeable difference is the smallest change you can detect, measured as the difference at which you say different and same equally often. Weber's law says that difference grows with the size of the starting stimulus, roughly in proportion. That pattern appears in almost every sensory dimension, which is why the same added amount is easy to notice against a small baseline and hard against a large one.

When evidence on a trial exceeds an internal criterion, you report a signal. Saying yes is a hit on signal trials and a false alarm on noise trials; saying no is a miss or a correct rejection. Because a falling hit rate can come from lower sensitivity or from a stricter criterion, you need the false alarms too to tell the two apart.

Sensory neurons adjust their sensitivity to recent stimulation, so a steady or repeated input, such as a lingering odor, comes to feel weaker. Adaptation appears in every sense and at several stages, from the receptors to the brain, and it lets neurons cope with inputs whose level can swing more than a hundredfold. Exactly what it is for, and how it works, is still debated.

Light passes through the cornea, which bends it; the iris sets how much enters through the pupil, and the lens works with the cornea to focus it on the retina, where photoreceptors turn it into electrical signals. A surface's color depends on which wavelengths it reflects, but not one to one: physically different surfaces, called metamers, can look identical under the same light.

In dim light only rods are active; in bright light rods saturate and three classes of cones, with overlapping wavelength sensitivities, provide trichromatic color. Cones dominate the fovea, the center of gaze, while rods are most common just around it. At intermediate light levels both work at once, and the range of colors you can see shrinks.

Trichromatic coding happens first: three cone types with overlapping sensitivities. Opponent coding comes next. Midget ganglion cells in the retina compare long- and medium-wavelength cone input to signal red versus green, and another cell class sets short-wavelength cones against the other two. So the two theories are not rivals; they describe successive stages of the same pathway.

Sound waves travel down the ear canal and vibrate the eardrum. Three tiny middle-ear bones, the malleus, incus and stapes, amplify the vibrations and pass them to the fluid-filled cochlea in the inner ear. There a traveling wave runs along the basilar membrane, the hair cells riding it convert the motion into electrical signals, and the auditory nerve carries those signals to the brain.

Place coding reads pitch from where along the cochlea hair cells respond, higher pitches near its wide end. Temporal coding reads it from auditory-nerve spikes that keep time with the wave, called phase locking. Many fibers top out near 200 spikes a second, so in Wever's volley theory the population, not one fiber, carries the timing. Phase locking fades at high frequencies, with estimated upper limits from 1,500 to 10,000 hertz.

Taste cells respond to at least five basic qualities: sweet, sour, bitter, salty and umami. They are scattered across the tongue, so the idea of separate taste zones is a myth. Each olfactory neuron carries one odor receptor and connects directly to the brain, and a molecule triggers a combination of receptors whose pattern the brain reads as a smell. Flavor blends both senses.

Gate control theory, proposed by Melzack and Wall in 1965, holds that circuits in the spinal cord's dorsal horn let strong touch input damp pain from the same spot, which is why rubbing a fresh injury helps. Balance relies on the inner ear: semicircular canals sense head rotation, while the utricle and saccule sense head position relative to gravity, combined with signals from the eyes and from muscles and touch sensors in the legs.

Bottom-up processing builds a percept from simple features, stage by stage from the retina to higher visual areas. Top-down processing brings expectations to bear on the same input. Feature-building alone cannot explain many effects, and one still-speculative view treats perceptual wholes as the brain's predictions, used to settle ambiguous input, with higher areas modulating early visual cortex.

Grouping principles such as proximity, similarity and good continuation are treated as special cases of one tendency: perception settles on the simplest, best overall organization. By similarity, for example, uniformly colored parts group together. When several principles agree, groups segment sharply; when they conflict, one principle may win or the percept may switch back and forth between organizations.

You estimate distance from several cues at once. Binocular disparity underlies stereo vision, while monocular cues such as retinal image size, motion parallax and looming work with one eye and are closely tied to movement. When mice had one eye closed, they still jumped gaps accurately by moving their heads more, a sign that monocular cues can stand in when stereo vision is lost.

Constancy lets you see a familiar object as keeping the same shape, size and color despite changes in perspective, distance, lighting and the size of its image on the retina. The same machinery can mislead. Illusions such as the Ponzo and Muller-Lyer are explained by size constancy, which means context can manipulate how large something looks.

Inattentional blindness is failing to notice a fully visible but unexpected object while your attention is on a task; about half of viewers counting ball passes miss a person in a gorilla suit. In change blindness, people often need ten to twenty alternations of two pictures to spot a large difference. Inattention may also explain why experts, such as radiologists, miss findings in plain view.

An unconditioned stimulus, such as food in the mouth, triggers an unconditioned response without any learning. A neutral cue repeatedly paired with it becomes a conditioned stimulus, and it then produces a conditioned response on its own. To label a new case, ask which stimulus caused the reaction before any pairing; the cue has to start out neutral and must not already produce that reaction.

Order matters during acquisition. In classic pigeon studies, a lit key followed by food came to draw pecks, while control groups without that forward pairing, including food delivered before the light, did not peck reliably. Timing alone is not the full story, though: modern research treats the plain pairing account as too simple, because learning depends on how well the cue predicts the outcome.

Extinction is not unlearning. Presenting the cue without its outcome builds a second, context-bound association, cue predicts nothing, that competes with the original, which stays intact. That is why the response can come back: after time passes (spontaneous recovery), in a new setting (renewal), or when the outcome occurs again (reinstatement).

After conditioning, stimuli resembling the conditioned cue also draw a response, and the response weakens the less they resemble it, along a generalization gradient. Discrimination is learning to respond to the predictive cue but not to a similar cue that predicts nothing. Overgeneralization, a reduced ability to tell predictive from non-predictive cues, is studied in fear and in appetite.

A flavor followed by sickness becomes avoided, the Garcia effect. It bends the usual rules: the taste-illness gap can be hours, with records of 24 hours, and the aversion is hard to extinguish. The outcome also picks the cue. In Garcia and Koelling's 1966 study, rats made ill after drinking avoided the flavor but not the lights and noise that came with it, while shocked rats did the reverse.

Watson and Rayner's 1920 Little Albert study reported conditioning a fear response in one young child, an early attempt to show a laboratory effect in a more natural setting. Watson's student Mary Cover Jones then used exposure, close to Pavlovian extinction, to reduce conditioned fears in children, the line that led to systematic desensitization. The same logic appears in cue exposure therapy, which reduces cue-triggered food cravings.

Thorndike's cats escaped puzzle boxes faster with practice as the required response became more likely. His law of effect says that responses closely followed by satisfaction become more firmly connected to the situation, so they recur when it recurs. That wording makes it a theory, the learning of a situation-response link, rather than a mere summary of faster escapes.

Reinforcement increases a behavior and punishment decreases it; positive means something is added and negative means something is taken away. Paying a child for good grades is positive reinforcement, and excusing a child from a dull outing for being kind is negative reinforcement. A spanking is positive punishment, and a time out, which removes something pleasant, is negative punishment. Label the effect first, then the add or remove.

Negative reinforcement increases a behavior by withdrawing something, usually something unpleasant, so it is the opposite of punishment in its effect. The "negative" refers only to removal. A child who is let off a boring outing for being kind to siblings becomes kinder. Escape and relief can also feel like gaining safety, which is one reason the categories blur in practice.

A ratio schedule pays off after an amount of activity, an interval schedule after a passage of time; each can be fixed or variable. Fixed schedules produce a pause before work begins again, and the pause grows with the size of the ratio or interval. Variable-ratio schedules, like hunting or gambling, keep activity high, but a ratio set too steep stops responding altogether, called ratio strain.

Behavior learned when only some responses pay off extinguishes less readily than behavior rewarded every time, the partial reinforcement extinction effect. Variable-ratio payoffs, as in gambling, are a case of this. One leading explanation, Amsel's frustration theory, holds that missing an expected reward energizes responding. The effect is not universal: within-subject studies have sometimes found the reverse.

Shaping builds a target behavior by reinforcing successive approximations to it; the bar for reinforcement narrows over time, so responses must look ever more like the target to pay off. Chaining, Skinner's term, links shorter behaviors into a sequence that ends in a primary reward, with each step held in place by conditioned reinforcement from the next.

In Bandura's Bobo doll studies, children who watched an adult attack an inflated doll tended to copy the aggression, with no reward needed. Bandura described four steps: attention to the model, retention of what was seen, reproduction of it in action, and motivation to act. Consequences feed the last step: seeing the model punished made imitation less likely, while a rewarded or unpunished model was copied more.

In latent learning experiments, animals that first explore a maze without reward learn faster once a reward appears, showing they learned the layout before their performance showed it. Tolman took this as evidence of a cognitive map. His famous shortcut test in the sunburst maze has replicated poorly, though: across about a dozen replications, only 17% of experiments showed shortcutting.

Remembering a fact or an event takes three steps: encoding the content, storing it, and later retrieving it. Asking which step failed separates cases that look alike. In an alcohol blackout, the gap arises because drinking blocks the transfer of memories into long-term storage, so the memory was never kept, while a memory distorted by later suggestion was stored but rebuilt wrongly at retrieval.

Sensory memory is the first store in the classic modal model. It holds a very large amount of incoming information, but only for a few hundred milliseconds, and its visual part is called iconic memory and its auditory part echoic memory. Only a subset of what it holds moves on to short-term memory, which is far more limited in capacity.

Short-term memory holds a small amount of information for seconds to minutes; Miller's classic estimate was about seven items. Working memory recasts that store as an active workspace where you manipulate a limited number of items, as when you rearrange numbers in your head. The difference is the emphasis: a holding area for what you just saw versus a place where thinking happens.

Miller proposed that short-term memory's limit of roughly seven items can be overcome by recoding information into chunks, units that each carry more. A string of digits held as a few familiar dates is one example. Chunks are mental representations, so building and unpacking them takes mental procedures for encoding and decoding, not just more storage space.

Processing an item's meaning, such as judging what a word means, leads to better recognition than shallow processing, such as judging what it rhymes with, both immediately and after a delay. A recent set of three experiments tested what the framework assumed: deep processing gave a head start, not slower forgetting, because the forgetting curves ran parallel.

Declarative memory is content you can, in principle, bring to mind: episodic memory for events from your own past and semantic memory for facts. Nondeclarative memory is defined by what it lacks, conscious content, and its forms share little else. It includes procedural memory, knowing how to ride a bicycle, and priming, where earlier exposure speeds a response to a related stimulus.

Context-dependent memory is better recall when the setting at test matches the setting at learning; an internal state, such as being on or off a drug, can act as context too. The classic demonstration had divers learn words on land or underwater, with recall best in the matching place. A 2021 replication with sixteen divers did not find that advantage, and the dive result has not often been replicated successfully.

Interference runs two ways: in retroactive interference later learning disrupts memory for earlier material, and in proactive interference earlier learning gets in the way of new material. Decay over time is studied as another passive cause. Forgetting can also be intentional: told to forget an outdated list, people recall it worse and the new list better, a goal-directed effect thought to rely on inhibition.

Remembering rebuilds an event from stored pieces rather than replaying it. In Loftus and Palmer's study, people who watched a car crash and were asked how fast the cars "smashed" gave higher speed estimates than those asked about cars that "hit", and later falsely recalled broken glass. Misleading post-event information is a likely by-product of reconstruction, though some cognitive abilities reduce susceptibility.

Under clean testing conditions, a double-blind procedure, a fair lineup and confidence recorded right after the first identification, high eyewitness confidence signals high accuracy. That is why confidence should be documented at the first test, before anything can contaminate it. The memory itself can also be distorted by post-event misinformation, which has cast doubt on eyewitness evidence in court.

Strategies with strong research support include retrieval practice, bringing material back to mind from memory, spaced practice, spreading study sessions out over time, and interleaving, switching between topics as you study. Retrieval takes effort, and that effort strengthens the memory. Spacing can make you feel less confident than cramming, because it is harder, yet that desirable difficulty is what helps long-term learning.

Prototype theory says something counts as a member of a concept when it has enough of the concept’s typical features, not when it meets a definition, which is why an apple feels like a better example of fruit than a plum. Its weak spot is combination: the prototype for “pet fish” is brightly colored, a feature that neither “pet” nor “fish” supplies.

A schema is a framework abstracted from many similar experiences, and a script is a schema for a routine event, such as being seated, ordering and eating at a restaurant. It gives new details something to attach to and a sequence to step through when you recall them. The cost is that a very strong schema can produce false memories that fit it too neatly.

Unlike an algorithm, a heuristic is a rule of thumb: it searches far more efficiently than exhaustive trial and error, but it does not guarantee a solution. The trade is accuracy for effort, and the errors heuristics produce are systematic rather than random, which is why researchers can name and predict them as biases.

Availability judges how likely something is by how easily examples come to mind, so a vivid, recent image can make a rare event feel more probable than the statistics say. Representativeness judges by similarity to a type, which leads people to neglect base rates and sample size and to expect chance to correct itself, as in the gambler’s fallacy.

Anchoring starts an estimate from an initial value and adjusts from there, but the adjustment is often too small, even when the anchor is irrelevant. Framing changes choices without changing facts: people tend to avoid risk when outcomes are described as gains and to seek risk when the same outcomes are described as losses. A meta-analysis found the frame explains most of that shift, though no standard choice model fully captures it.

Confirmation bias is the tendency to favor information that fits what you already believe and to discount evidence against it, and it shapes how you interpret events and even what you later recall. One modeling paper argues the bias can be adaptive when paired with good self-monitoring, so you discount noise but still seek new evidence once you realize you are wrong.

Functional fixedness locks you onto an object’s usual use: in Duncker’s candle problem, people miss that the thumbtack box can be emptied and tacked to the wall to hold the candle. Mental set does the same to methods, as in Luchins’s water-jar studies: a familiar solution from earlier problems keeps returning and crowds out alternatives, even when it no longer fits.

Language is analyzed in levels. Phonemes are the smallest sound units that change a word, so “cat” has three; morphemes are the basic units of meaning, some free like “cat” and some bound like “un-” or “-ing,” which only attach to a base. Syntax concerns the form of sentences, and semantics concerns how expressions relate to what they denote.

Babies babble speech-like sounds such as p, b and m by about 4 to 6 months, and the first three years are the most intensive period for learning language. Deaf people first exposed to a sign language late show weaker grammar the later they started, evidence for a critical period. Nativists argue the input children hear is too poor to explain learning without an innate language system; empiricists dispute that.

Scores on very different mental tests correlate positively, a pattern called the positive manifold that Spearman first described in 1904. He explained it with a single general factor, g. Researchers broadly accept that g summarizes those correlations, but what g actually is, mental speed, working memory capacity or something else, is still debated.

Fluid intelligence is reasoning through novel problems without relying on what you already know; crystallized intelligence is accumulated learning and world knowledge. They age differently: fluid measures tend to decline in later life while crystallized measures tend to stay stable. Generational score gains have also been about twice as large on fluid measures.

Gardner proposed several independent intelligences, including linguistic, musical, bodily-kinesthetic and interpersonal, and Sternberg proposed analytical, creative and practical intelligence. The main objection to Gardner is evidential: the separate brain systems his theory predicts have not been found, and many researchers now classify the theory as a neuromyth.

An IQ is not a count of right answers. It expresses how far a performance sits from that of same-age peers in the test’s normative sample, on a scale with a mean of 100 and a standard deviation of 15. Because it is relative, a score near a test’s floor carries little information, since the test cannot distinguish among very low performances.

Because IQ is scored against a norm group, norms age. Average scores rose about two to four points per decade over much of the twentieth century, the Flynn effect, so old norms inflate scores; tests are re-normed on census-matched samples, and one guideline for diagnosing intellectual disability subtracts 0.3 points per year since norming. Gains have since slowed, stalled or reversed in some countries.

A test that selects groups at different rates shows adverse impact, but a gap alone does not prove the test is biased; general cognitive tests produce larger group differences than some other valid predictors. Predictive unfairness means a group performs better or worse on the job than its scores indicated, compared with other groups, and adverse impact is the trigger to check for it.

Stereotype threat is the proposal that people underperform when a negative stereotype about their group is made salient, perhaps because it burdens mental resources; Steele and Aronson’s verbal-test study is the classic case. It is now contested: a meta-analysis of 47 lab studies of gender and math warned that publication bias may have seriously distorted the estimate, and a preregistered field replication of one proposed fix found no effect.

IQ predicts school performance moderately: a meta-analysis of over 42,000 people put the correlation near 0.37, and about 0.5 is widely cited. One critical review notes the school link may be partly circular, since tests and school measures share content, and argues the often-quoted job correlation of about 0.5 rests on heavy statistical corrections to old data of uncertain quality.

Gene-environment interaction means a trait’s expression depends on genes, environments and how they act on each other, so the influence of a gene can change with the environment, and the reverse. Asking “nature or nurture?” misses this, and a better grasp of interaction has been linked to less deterministic thinking about health outcomes.

A cross-sectional study compares different age groups at one time, so an age difference may really be a cohort difference between generations; such studies place the start of cognitive decline in the thirties. A longitudinal study follows the same people, separating aging from cohort, but it suffers from selective dropout and from practice effects of repeated testing.

Most birth defects arise in the first three months of pregnancy, while the organs are forming; the fetal period, when much of the brain’s neuron production occurs, begins after about eight weeks. Timing shapes risk, but no stage is fully protected: some defects arise later as tissues keep developing, and there is no known safe time to drink alcohol during pregnancy.

Piaget proposed four stages. Sensorimotor infants (about 0 to 2) gain object permanence, an inner representation of an object. Preoperational thinking (about 2 to 7) is egocentric and fixed on one feature. In concrete operations, from about 6 or 7, grasping that a change can be reversed lets children conserve, seeing a quantity stay the same when only its arrangement changes. Formal operations (from about 11) bring abstract thought.

Vygotsky’s zone of proximal development is the range between a learner’s unaided performance and what they can do with help from a teacher or a more capable peer. Scaffolding is the help itself: support matched to the learner’s current level so they can finish a task they could not do alone, withdrawn as competence grows. One critic argued the idea invites overoptimism that any child can be taught anything.

Ainsworth’s Strange Situation separates an infant from a caregiver and then reunites them. Most infants in her sample were secure, using the caregiver as a base for exploring and a comfort on return; a minority were avoidant, ignoring the caregiver at reunion, or resistant, seeking contact without being soothed. Main later added disorganized. Meta-analyses link early attachment only modestly to later peer competence and behavior.

Crossing demandingness with responsiveness gives four styles: authoritative is high on both, authoritarian is demanding but not responsive, indulgent (permissive) is responsive but not demanding, and neglectful (uninvolved) is low on both. Much of the evidence linking styles to child outcomes is cross-sectional, so longitudinal work is still needed to settle the direction of influence.

Erikson described eight psychosocial tasks, each a tension between two poles, from trust versus mistrust in infancy through identity versus role confusion to generativity versus stagnation in midlife and integrity versus despair at the end. Because his account carries development into old age, it is credited as the first life-span theory.

Kohlberg organized moral development into three levels, preconventional, conventional and postconventional, each split into two stages, with reasoning from universal principles of justice at the top. Gilligan argued the scheme ranked a morality of rights above a morality of responsibility and care, instead of treating them as different, and so described many women as stuck short of full development.

The prefrontal cortex, which supports planning, prioritizing and good decisions, is one of the last brain regions to mature, and the brain keeps developing into the mid-to-late twenties. The dual-systems model proposes that an earlier-maturing reward system outpaces this later control system, which is thought to tilt teens toward risk, especially around peers; it is a model still being tested.

Taking longer to learn something new or occasionally misplacing things is usually normal age-related forgetfulness. Memory problems that make everyday tasks hard, such as repeating the same questions or getting lost in familiar places, point beyond normal aging. Socioemotional selectivity theory adds that as people’s sense of future time changes, their goals tilt toward emotion and close relationships.

Drive-reduction theory treated motivation as an unpleasant state, like hunger or thirst, that pushes you to act, with relief as the reward. That fits returning the body to its set points, but not the need for stimulation: in one study people gave themselves painful shocks rather than sit with nothing to do. Incentive accounts, in which rewards pull rather than drives push, largely replaced it.

In Yerkes and Dodson’s mouse experiments, performance rose steadily with shock intensity on an easy discrimination but peaked at an intermediate level on a hard one. So the law says arousal and performance can relate in a straight line or a curve depending on task difficulty, though it is often misremembered as one curve for every task.

Intrinsic motivation is doing something out of interest and enjoyment; extrinsic goals depend on outside outcomes such as others’ admiration. The undermining effect is a drop in intrinsic motivation after a performance-based reward is added, and it can only appear if there was interest to undermine. It is not universal: a study of children with and without ADHD found no drop, even for an interesting task.

Homeostatic eating is controlled mainly by the hypothalamus, which reads signals from the gut, fat tissue and the vagus nerve. Leptin, made by fat cells, suppresses appetite, and ghrelin, made mostly by the stomach, stimulates it; in one cohort, short sleep went with lower leptin and higher ghrelin. Hedonic eating can override the system, driving people to eat for pleasure regardless of their metabolic state.

Maslow ordered needs from physiological through safety, belonging and esteem up to self-actualization, to be satisfied one level at a time. The pyramid is intuitive, but a review of 13 studies found little evidence that needs emerge in a fixed order, and later research supports the range of needs Maslow named but not their hierarchy.

An episode of fear has several parts: an appraisal of danger, bodily changes such as a faster heart rate, an unpleasant feeling, an expression such as raised upper eyelids, an urge to flee, and narrowed attention. Theories of emotion largely differ over which of these parts are essential, and there is still no agreed definition.

James argued that the bodily response comes first and your feeling of it is the emotion: you are afraid because you tremble. Cannon objected that the visceral changes in fear and anger are too alike to tell emotions apart, and with Bard proposed that the felt emotion and bodily arousal occur at the same time and independently, giving the thalamus a central role.

Two-factor theory holds that arousal is much the same across emotions, so what you feel depends on how you label it from the situation. In Schachter and Singer’s experiment, some men injected with adrenaline but not told its effects took on a planted participant’s cheerful or angry mood, while correctly informed men did not. Critics say such views risk reducing emotion to verbal labeling.

Ekman’s basic emotion theory holds that a few emotions produce the same facial expressions in every culture, with culture-specific display rules able to mask them. Critics fault the methods, such as forced-choice answers that inflate agreement and actors’ posed faces, and constructionists argue there is no one-to-one match between an emotion and any facial or bodily signature.

The facial feedback hypothesis says your expression feeds back into what you feel. In Strack’s 1988 study, people holding a pen in their teeth, which engages a smiling muscle, rated cartoons more positively than people holding it in their lips. A multi-lab replication with 1,894 participants failed to find the effect; one exploratory analysis suggests it depends on social context, such as an experimenter’s presence.

The person-situation debate asked whether traits predict behavior across situations, after studies showed small situational factors swaying conduct: hurried passersby helped a person in distress far less often (10 versus 63 percent). Today mainstream researchers are interactionists, treating behavior as a joint function of person and situation; the open question is how to describe stable individual differences, not whether they exist.

In Freud’s model the id is the seat of instinctual desires, run by the pleasure principle; the superego is an internalized moral censor built largely from parental values; and the ego mediates between them and outside reality. Defense mechanisms are the ego’s unconscious ways of easing that conflict. A classic objection from philosophy of science is that the theory was framed so it could only be confirmed, never refuted.

Rogers held that people share one basic motive, an actualizing tendency toward growth that unfolds on its own in the right social climate, and he traced psychological problems to incongruence. The climate he described is experienced as unconditional, positively regarding, empathic and genuine; without it, growth is thwarted and people develop in less fully functioning ways.

The trait approach grew from the lexical hypothesis that important personality differences are encoded in everyday language. Allport and Odbert found about 4,500 trait-describing words in the dictionary; Cattell factor-analyzed a reduced list and settled on 16 factors, and later researchers with more computing power repeatedly found five, the model later named the Big Five.

The Big Five are openness, conscientiousness, extraversion, agreeableness and neuroticism, often scored in reverse as emotional stability. In school, conscientiousness predicts grades most consistently, and one meta-analysis found its effect rivaled cognitive ability’s; it also tracks how far people go in education. Openness shows small positive links, and the other three relate to school outcomes more weakly and inconsistently.

Social-cognitive theory’s organizing principle, reciprocal determinism, is a continuous interaction among person, environment and behavior, each shaping the others. Rotter’s locus of control describes one person factor: with an internal locus you see events as following from your own actions and traits; with an external locus you see them as luck, chance, fate or forces beyond you.

Objective inventories such as the MMPI-2-RF score fixed self-report items and add validity scales that flag underreporting and overreporting; in one study of parents in custody disputes, about 44 percent of profiles suggested underreporting. Projective tests such as the Rorschach ask open questions like “what might this be?” and are criticized for weak validity; their early scoring systems resisted comparison until Exner unified them in 1974.

The Barnum effect is accepting vague, general personality descriptions as accurate portraits of yourself. It runs on subjective validation, where a description feels right because it is personally meaningful, and on flattery, since people readily believe what makes them look good. Positivity and a scientific-looking source or procedure also shape it, which is how low-validity personality tests keep being accepted as accurate.

Situationism stresses that minor, seemingly irrelevant features of a situation often sway behavior more than character does. People who had just found a dime in a payphone helped a stranger who dropped papers far more often (88 versus 4 percent). Treat any single study with care, though: social psychology has had prominent replication failures, and conclusions should not rest on a few studies.

Attribution theory asks whether you explain an action internally, by the person, or externally, by the situation. Kelley’s covariation model says people pick an internal cause when the actor has acted this way before, acts this way in many contexts, and acts differently from others. Real perceivers often stray from that logic, and they usually explain behavior by specific motives, not just person versus situation.

The fundamental attribution error, or correspondence bias, is choosing a person’s disposition as the cause of their behavior even when an obvious external cause is present. One account is effort: perceivers generate internal explanations automatically and weigh the situation only when they are both motivated and able. Culture matters too, since Western samples tend to favor internal causes.

In the actor-observer bias, people explain others’ actions by stable characteristics but see their own actions as also shaped by circumstances. The self-serving bias is taking responsibility for success while blaming others or circumstances for failure; it expresses positive beliefs about the self, and many researchers see it as part of healthy psychological functioning.

The belief in a just world, proposed by Lerner, is the assumption that people get what they deserve, which keeps the world feeling stable, orderly and safe. When an injustice cannot be repaired, people often protect the belief by blaming the victim instead of helping, and stronger just-world belief correlates with more victim derogation.

Cognitive dissonance theory says inconsistency between attitudes and behavior is resolved, often, by shifting the attitude toward the behavior. Festinger and Carlsmith varied the pay for a discrepant act; because a good reason for the act is a consistent cognition that lowers dissonance, smaller payments should yield more attitude change. Critics question the classic methods, and as of 2019 large replication projects had not yet tested it.

In the elaboration likelihood model, the central route means weighing a message on its merits, while the peripheral route relies on simpler cues, such as how credible the source seems or how polished the message looks. Which route a message travels depends on whether the audience is motivated and able to process the arguments; when they are not, cues do the persuading.

In Asch’s line studies a lone participant heard several confederates give the same wrong answer before judging, and going along cost nothing. Normative influence is conforming to be accepted; informational influence is taking others’ answers as evidence of what is correct. A 2023 replication found wrong answers on 33 percent of judgments, matching Asch, and paying for accuracy cut that only to 25 percent.

In Milgram’s best-known version every participant gave shocks of at least 300 volts and 65 percent went to 450, but across about 30 variants obedience ranged from 0 to nearly 100 percent, and most participants overall stopped short. Closeness to the learner lowered it, and people tended to continue when told the experiment required it but usually refused a direct order. Replications have had mixed results, and the ethics are still debated.

Foot-in-the-door asks a small favor first, then a larger related one, and draws its force from commitment to that first yes. Door-in-the-face runs the other way: a large request likely to be refused, then the real, smaller one, which feels like a concession that deserves one in return. Neither is reliable everywhere; in one field study, door-in-the-face produced less compliance than a control condition.

Janis defined groupthink as the thinking of members deeply involved in a cohesive in-group, whose push for unanimity overrides realistic appraisal of alternatives. Its signs include illusions of invulnerability and unanimity, collective rationalization, self-censorship and belief in the group’s inherent morality. Stress and extreme situations may strengthen the pull toward conformity and groupthink.

Zajonc’s drive theory offers one explanation of social facilitation: others’ presence raises arousal, which strengthens your dominant response, so a well-learned task improves while an unlearned one gets worse. Social loafing is lower individual effort on a team task than alone, strongest when contributions cannot be identified or evaluated, the task seems low in value, or members’ inputs are redundant.

Deindividuation is the loss of individuality, personal responsibility and self-awareness that group membership can bring, though it may also feed affiliation and prosocial feeling toward fellow members. Group polarization is a like-minded group’s views growing more extreme as members mainly talk to each other, as in online echo chambers, where dissenting information seems mostly ignored and may even deepen the split.

Prejudice has cognitive, affective and behavioral components, which act jointly, and emotions toward a group may predict discriminatory behavior more directly than beliefs about it. Favoring your own group is a strong default: it appears even in minimal groups formed on trivial grounds, such as preferring paintings by Klee or Kandinsky.

The bystander effect is the finding that people are less likely to help as more bystanders are present, and the effect grows with their number. The main explanations are diffusion of responsibility, where no one takes on the duty to act, and pluralistic ignorance. One classic study found the effect only when the situation was ambiguous, and helping is likelier when the bystander shares an identity with the victim.

Industrial-organizational psychologists apply psychological research and methods to workplace problems, studying productivity, employee testing and selection, and organizational dynamics to improve work-life quality. The closely related field of human factors treats people, tasks, tools, organization and environment as one system and works to improve performance within it.

Job analysis examines the tasks a job involves, the competencies needed to perform them, and the link between the two. It is the foundation of selection because it shows which requirements are genuinely job-related, which is also what makes a selection procedure legally defensible, and its data feed training, classification and performance appraisal.

General cognitive ability tests predict job performance and training success across many jobs, and better as jobs grow more complex. Structured interviews, which ask every candidate the same questions in the same order and score answers on one scale, and work samples that mirror real tasks also relate highly to performance. Personality tests are valid but tend to predict less well.

Ratings carry predictable distortions. Halo is letting an overall impression color separate judgments; leniency is scoring higher than performance deserves, and severity the reverse; central tendency is avoiding the extremes and squeezing everyone toward the middle. The usual countermeasures are training raters to recognize these errors and to share a common frame of reference for standards.

Expectancy theory says effort depends on how a person judges the link between effort and outcomes, built from expectancy, instrumentality and valence, the value placed on the outcome. Goal-setting theory says specific goals focus effort on essential tasks and challenging goals push people to work harder and find better strategies, though goals seen as unattainable breed frustration.

Human factors engineering treats work as a system of people, tasks, tools, organization and environment, and improves performance by redesigning parts of that system. In one observational study of health workers removing protective gear, failures split into violations, procedural mistakes and slips, so the authors concluded that no single fix, such as training alone, would be enough.

A stressor is the outside cause, such as a heavy homework load or an illness, and it may be brief or recur over a long time; stress is the physical or mental response to it. Keeping them apart matters because the response is strongly shaped by how the person appraises the stressor, so the same event can produce very different reactions.

In Lazarus and Folkman’s transactional model, stress comes from the person and environment together. Primary appraisal weighs how relevant a stressor is; secondary appraisal weighs whether your resources can overcome it. Both steer the coping strategies chosen, and coping in turn affects the immediate stress response and longer-term health and well-being.

Selye noticed that patients with different diseases shared similar symptoms and proposed a general, nonspecific stress response in three stages: alarm, with acute signs; resistance, as those signs fade; and exhaustion, when the reaction returns or the organism collapses. Later work revised the idea into allostasis, which lets set points shift and accounts for anticipatory responses to stressors.

Psychoneuroimmunology finds that the duration and intensity of stress matter: brief, mild-to-moderate stress can enhance immune function, while chronic stress with prolonged cortisol can weaken antibody responses and vaccine efficacy. Effects on serious disease are less certain; whether chronic stress is linked to cancer is not clear, partly because stress also shifts behaviors such as smoking and drinking.

Problem-focused coping tries to manage or change the problem causing distress; emotion-focused coping regulates the emotional response to it. The goodness-of-fit idea pairs problem-focused strategies with controllable situations and emotion-focused ones with uncontrollable situations, but findings conflict and many strategies serve both functions, so the two-way split is an oversimplification.

The buffering hypothesis says relationships supply informational, emotional or tangible resources that soften the harm stressors do to health; social support means the real or perceived availability of those resources. More broadly, a meta-analysis of 148 studies associated stronger social relationships with about 50 percent higher odds of survival, a link comparable to risks such as smoking.

Burnout is a syndrome of emotional exhaustion, cynicism and reduced personal accomplishment that arises from chronic stress, often in demanding work settings. Whether it is distinct from depression is debated: one meta-analysis found them correlated at about 0.52 yet called them different constructs, while a large physician study argued they overlap and that the three-part model is deficient.

Accounts of mental disorder lean on several marks: distress or suffering, disability or dysfunction, and departure from statistical norms. None works alone: being unusual is not being disordered, and ordinary grief can resemble depression. Wakefield’s harmful-dysfunction account therefore requires two things at once, harm judged by the person’s culture and an internal mechanism failing at its natural function.

The DSM and ICD largely follow a model in which each disorder is a bounded entity with its own symptom cluster. Critics propose classifying dimensionally instead, such as rating depression by severity, or as networks of interacting biological, psychological and social factors. NIMH’s research framework, RDoC, studies degrees of dysfunction in basic systems and is not meant to replace diagnosis.

The diathesis-stress model explains why the same stress triggers disorder in some people and not others: a latent vulnerability, often biological, is activated by stress, and symptoms appear once a threshold is crossed. Tests pairing polygenic risk scores with stressful life events have found such interactions for depression, but in one sample of 4,919 people the effect was only marginally significant and tiny.

Anxiety disorders, including generalized anxiety disorder, panic disorder, social anxiety disorder and phobias, involve fear or worry that does not go away, appears in many situations and can grow over time. In cognitive models, avoidance keeps them going: by staying away from what is feared, a person never gathers the evidence that the feared outcome does not happen, so the fear prediction survives.

In OCD, obsessions are uncontrollable, recurring thoughts and compulsions are repetitive, excessive behaviors. Compulsions bring no pleasure but give temporary relief from anxiety, and that relief reinforces them. Blocking the ritual during exposure lets a person find that the feared outcome does not follow, which is the logic of exposure with response prevention.

Strong reactions after trauma are common, and most people recover over time; PTSD is diagnosed when symptoms last and interfere with daily life. Its symptoms include re-experiencing, avoidance of reminders, heightened arousal and reactivity, and changes in thinking and mood. Risk rises with earlier trauma and with little social support afterward, while seeking support is a resilience factor.

Major depression means depressed mood or loss of interest most of the time for at least two weeks, interfering with daily life. Persistent depressive disorder is milder but lasts much longer, usually two years or more. Bipolar I involves manic episodes; bipolar II pairs depressive episodes with milder hypomanic ones; cyclothymia involves recurring mood symptoms too mild or brief to count as episodes.

The attributional version of learned helplessness says people prone to depression explain bad events by internal, stable and global causes. Beck’s cognitive theory adds the triad of negative views of the self, the world and the future. The popular idea that depression is caused by low serotonin overreached: a large umbrella review found no consistent evidence for it, though antidepressants may act in other ways.

Schizophrenia symptoms fall into three groups: psychotic, such as hallucinations and delusions; negative, such as lost motivation and difficulty showing emotion; and cognitive, problems with attention and memory that strongly predict daily functioning. The revised dopamine hypothesis ties psychotic symptoms to excess subcortical dopamine and negative ones to reduced prefrontal activity, while subtle brain differences may begin before birth.

Personality disorders are long-standing patterns of thinking and behaving that stay fixed across situations and lead to distress or impairment; a US survey from 2001 to 2003 put past-year prevalence at 9.1 percent. They are grouped into three clusters, odd or eccentric, dramatic or erratic, and anxious or fearful, but heavy overlap between categories led to a dimensional alternative model for further study.

ADHD and autism are neurodevelopmental conditions. ADHD involves inattention, hyperactivity or impulsivity that must begin before age 12 and often continues into adulthood; signs of autism generally appear in the first two years. Both vary widely, which is why autism is called a spectrum with a range of needs, strengths and challenges, and supports can improve daily functioning.

Psychoanalysis asks patients to free-associate, saying whatever comes to mind even if it seems irrelevant or nonsensical, and dreams are explored the same way. The analyst listens for links to unconscious conflicts, and the resistances and transference that arise become grounds for interpretation. The aim, in Freud’s phrase, is to remember, repeat and work through what was repressed.

Behavior therapies apply learning principles directly. Systematic desensitization pairs relaxation training with a graded hierarchy of feared situations, faced in imagination or in reality. Exposure therapy has a person tolerate feared items in brief, supported doses until the fear declines, and contingency management gives material incentives for behavior change that is objectively verified.

Cognitive behavioral therapy teaches people to notice unhelpful thoughts, such as a low opinion of their own abilities, question them, see how they shape feelings and behavior, and change self-defeating patterns. For depression, psychotherapy effects are comparable to antidepressants’, with evidence that CBT lasts longer by preventing relapse; whether it beats other psychotherapies is still argued.

Rogers’s client-centered therapy is non-directive: instead of steering the client, the therapist offers a relationship marked by unconditional positive regard, empathy and genuineness, trusting the client’s own capacity for growth. Rogers held that providing this climate is necessary and sufficient for constructive personality change, so the attitudes matter more than any technique.

Drug classes are grouped by target symptoms: antidepressants such as SSRIs and SNRIs for depression and often anxiety, anti-anxiety drugs for panic and extreme worry, stimulants for ADHD, antipsychotics for psychosis, and mood stabilizers for bipolar disorder. Electroconvulsive therapy uses an electric current to induce seizure activity and is usually considered only after other treatments have not helped.

The common factors view holds that what therapies share explains much of their benefit: therapist empathy and positive regard, a good working alliance, and positive expectations on both sides. Of these, the alliance is the most robust predictor of outcome. Specific techniques still matter as the credible framework the common factors depend on, and their relative weight remains debated.

Evidence-based practice rests on three components, a “three-legged stool”: research evidence, clinical expertise, and the client’s values, preferences and characteristics. “Empirically supported” is a narrower label from APA criteria that grade efficacy, with “efficacious and specific” at the top, and separately judge effectiveness, meaning how well a treatment carries over into everyday practice.

In 2022 about half of US adults with any mental illness, 50.6 percent, received treatment in the past year: roughly 57 percent of women versus 42 percent of men, and about 56 percent of White adults versus 36 to 40 percent of Asian, Black and Hispanic adults. Reviews tie lower use to cost, transport, scarce services, stigma and cultural beliefs; among young adults with untreated depression, cost was the reason given most often.