Operant Conditioning for AP Psychology: Terms, Studies, and Practice

Where operant conditioning sits on the AP exam

The College Board revised AP Psychology for the 2024–25 school year. In the revised Course and Exam Description (CED), learning is no longer a unit of its own: classical conditioning, operant conditioning, and social and cognitive factors in learning sit within Unit 3, "Development and Learning."^[1] The exam has a multiple-choice section and a free-response section; in the revised course the free-response questions ask you to analyze a description of a study and to argue from several sources of evidence rather than to define terms.^[1] Question counts, timing, and weighting are in the CED and can change between editions; check the current one rather than any summary, including this one.

You will rarely be asked to define negative reinforcement. You will be given a scenario — a parent, a rat, a paragraph from a study — and asked what kind of learning it shows or whether the design supports the conclusion. The CED is free and lists every term the exam can draw on; read it before you study and again after. Anything not in it is not required, and the last section here sorts out which pages on this site go past it.^[1]

The vocabulary, one line each

Definitions are functional, as the exam uses them: a reinforcer or punisher is defined by its effect on behavior, not by intent or feel. Each links to the page that treats it fully.

Consequences and the four quadrants

• Law of effect. Responses followed by satisfaction are strengthened, responses followed by discomfort weakened; Thorndike's wording is in the library.^[2][3]
• Operant conditioning. A voluntary behavior changes in frequency because of its consequences; Skinner's term.^[4]
• Reinforcement. Any consequence that makes the behavior it follows more frequent.
• Punishment. Any consequence that makes the behavior it follows less frequent.
• Positive and negative. Added and removed. Not good and bad.
• Positive reinforcement. Stimulus added, behavior up: a treat after a sit.
• Negative reinforcement. Stimulus removed, behavior up: buckling up silences the chime.
• Positive punishment. Stimulus added, behavior down: a scolding after jumping on the couch.
• Negative punishment. Stimulus removed, behavior down: losing the car keys after missing curfew.
• Extinction. The reinforcer stops following the behavior and the behavior declines. Not a quadrant: nothing is added or removed.
• Extinction burst. The brief rise in rate or intensity when reinforcement is first withheld.
• Spontaneous recovery. An extinguished response returns after a rest.
• Generalization. Responding to stimuli that resemble the training stimulus.
• Discrimination. Responding to the training stimulus but not to similar ones; the signal that reinforcement is available is a discriminative stimulus.

Reinforcers

• Primary reinforcer. Reinforcing without learning: food, water, warmth.
• Secondary (conditioned) reinforcer. Reinforcing through association with primary reinforcers: money, grades, praise.
• Token economy. Tokens earned for target behaviors are exchanged later for backup reinforcers.
• Premack principle. A more probable behavior reinforces a less probable one when made contingent on it: first homework, then the game.^[5]

Building and maintaining behavior

• Shaping. Reinforcing successive approximations to a target behavior.
• Chaining. Linking behaviors so each step cues the next and the last produces the reinforcer.
• Skinner box. An operant chamber with a lever or key, a reinforcer dispenser, and a cumulative recorder.^[4]
• Continuous reinforcement. Every response reinforced: fastest acquisition, fastest extinction.
• Partial (intermittent) reinforcement. Some responses reinforced: slower acquisition, far greater resistance to extinction. AP materials say partial; behavior analysts say intermittent.
• Fixed ratio (FR). Reinforcement after a set number of responses. Pattern: a pause after each reinforcer, then a fast run.^[6]
• Variable ratio (VR). After an unpredictable number of responses. Pattern: high and steady, hardest to extinguish.^[6]
• Fixed interval (FI). The first response after a set time. Pattern: the scallop, accelerating as the interval ends.^[6]
• Variable interval (VI). The first response after an unpredictable time. Pattern: moderate and steady.^[6]

Named phenomena

• Superstitious behavior. Strengthened by accidental, non-contingent reinforcement; Skinner's pigeons, fed every fifteen seconds whatever they did, developed stereotyped turning.^[7]
• Learned helplessness. After uncontrollable aversive events, an organism fails to escape when it can; Seligman and Maier's dogs.^[8]
• Instinctive drift. Trained behavior drifts toward species-typical behavior; the Brelands' raccoons "washed" the coins they were trained to deposit.^[9]
• Biological preparedness. Some associations are learned easily and others hardly at all; rats link taste with illness and light or sound with shock, not the reverse.^[10][11]

Contrast cases: cognitive and social learning

• Latent learning. Learning without reinforcement that appears when there is reason to use it; Tolman and Honzik's rats had a cognitive map before food was offered.^[12]
• Insight learning. A sudden, complete solution rather than gradual trial and error; Köhler's chimpanzees.^[13]
• Observational learning (modeling). Learning by watching a model; Bandura's Bobo doll. Vicarious reinforcement and punishment are consequences the observer sees the model receive.^[14]

Operant vs. classical conditioning, as the exam frames it

The exam treats these as the two kinds of associative learning and tests whether you can tell them apart. In classical conditioning two stimuli are paired and a reflex transfers from one to the other: Pavlov's dogs salivated to the signal that preceded food, and Watson and Rayner's infant cried at a white rat that had been paired with a loud noise.^[15][16] In operant conditioning a voluntary behavior is followed by a consequence and changes in frequency. Classical responses are elicited by what comes before; operant responses are emitted and changed by what comes after.

Extinction, spontaneous recovery, generalization, and discrimination occur in both, so a question can ask about any of them in either frame. Most real scenes contain both — a dog excited at the treat pouch, then sitting for the treat. The operant vs. classical conditioning page has the decision checklist and eight worked examples.

The four confusions the test writers count on

Most wrong answers come from one of four confusions, and each has a distractor written for it.

1. Negative reinforcement is not punishment

Reinforcement or punishment says what happened to the behavior. Positive or negative says what happened to the environment. Negative reinforcement is "behavior up, something removed." Nagging that stops when the room is clean negatively reinforces cleaning; nagging that starts at the sight of a mess positively punishes messiness. The same stimulus does opposite jobs.

2. Positive means added, even when what is added is unpleasant

"Positive punishment" is a contradiction only if positive means good. It means added: a shock, a scolding, an extra chore. And a consequence is a punisher only if the behavior decreases; a detention that does not reduce talking is not punishment, whatever it was called. Items exploit this with a "punishment" followed by more of the behavior.

3. Ratio vs. interval, and what "variable" does not tell you

Students learn that variable schedules resist extinction and then call anything unpredictable "variable." The first question is ratio or interval. If reinforcement depends on how many responses occur, it is ratio and responding faster pays sooner; if it depends on time, it is interval and responding faster changes nothing. A slot machine paying after an unpredictable number of pulls is variable ratio; a pool where fish bite after unpredictable stretches of time is variable interval. Only the gambler's rate matters, which is why the gambler pulls fast and the angler casts at a moderate pace.^[6]

4. Extinction is not punishment

Both reduce behavior. In punishment something happens after the response: an aversive is added or a reinforcer removed. In extinction nothing happens; the usual reinforcer simply stops. Tantrums once maintained by attention and now ignored are on extinction; a toy removed after each tantrum is negative punishment. Extinction also has two signatures punishment lacks: a burst at the start and spontaneous recovery after a rest.

The studies you should be able to name

Questions name the researcher and expect the term, or describe the procedure and expect the researcher.

The Bobo doll, latent learning, and insight rows are contrast cases: learning occurred without reinforcement, though reinforcement governs whether it is performed.

How to work a scenario question

Whether the scenario is a multiple-choice stem or a paragraph from a study, the same six moves name the process.

1. Find the behavior. What the organism does, as a verb. If two people are in the scene, decide whose behavior is being asked about.
2. Find the consequence. What happened immediately after, for the one behaving.
3. Added or removed? Added is positive; removed is negative.
4. Up or down? "More likely," "keeps doing it," and "continues" mean up; "stops" and "less often" mean down.
5. Name the quadrant. Added and up: positive reinforcement. Removed and up: negative reinforcement. Added and down: positive punishment. Removed and down: negative punishment.
6. Check that it is a quadrant at all. If the usual reinforcer simply stopped, it is extinction. If a reflex was elicited by a signal, it is classical. If the learner watched someone else, it is observational learning.

A worked case. Maya's brother screams whenever she takes the tablet; she gives it back to stop the noise, and he screams sooner next time. For the brother, the tablet is returned (added) and screaming goes up: positive reinforcement. For Maya, the screaming stops (removed) and handing it back becomes more likely: negative reinforcement. The A-B-C model page works this decomposition in detail.

Naming the schedule from the pattern

Two questions. Does reinforcement depend on a count of responses ("every tenth," "an unpredictable number of") or on time ("after a set time," "at varying intervals")? Count is ratio, time is interval. Is the requirement the same each time or does it vary? Then confirm with the pattern — pause-and-run, high and steady, the scallop, moderate and steady — because a question that describes pacing is handing you the answer. The schedules of reinforcement page has a simulator that draws each cumulative record live; once you have watched the scallop form you will not confuse it with the ratio pause.

When free-response material describes a study, read the method a sentence at a time and name each procedure as you meet it — the reinforcer, the schedule, the dependent variable — because a question about whether the conclusion follows usually turns on a procedure named correctly.

Practice set: ten items with answers

Work each item with the six moves before opening the answer. The examples page has fifty more, and the quiz gives instant explanations.

1. A student gets a sticker every time she turns in homework, and submissions increase. Quadrant and schedule?

   Answer

   Positive reinforcement (sticker added, behavior up) on continuous reinforcement, so expect fast extinction if the stickers stop.

2. A driver buckles up to silence the chime and, over time, buckles up faster. Quadrant?

   Answer

   Negative reinforcement: the chime is removed and the behavior increases. Not punishment, because nothing decreased.

3. A teenager comes home after curfew and loses driving privileges for a week; late arrivals become less frequent. Quadrant?

   Answer

   Negative punishment: a reinforcer is removed and the behavior decreases. Had late arrivals continued, it would not have been punishment, whatever the parents called it.

4. A rat gets a pellet after every tenth press; its cumulative record shows a flat stretch after each pellet, then a steep run. Schedule and pattern?

   Answer

   Fixed ratio 10. The flat stretch is the post-reinforcement pause and the steep run the high-rate burst: break-and-run.

5. A gambler keeps pulling a machine that pays after an unpredictable number of pulls; an angler keeps casting where fish strike after unpredictable stretches of time. Name each schedule and the phrase that decides it.

   Answer

   Variable ratio: "number of pulls" makes reinforcement depend on a count. Variable interval: "stretches of time" makes it depend on time. Both are unpredictable, so "variable" alone does not decide it.

6. A child's tantrums were maintained by attention. The parents now withhold it; tantrums briefly get louder, then decline. Name the process and the brief increase.

   Answer

   Extinction, not punishment: the reinforcer stopped, and nothing is added or removed after each tantrum. The increase is the extinction burst; removing a toy after each tantrum would be negative punishment.

7. A pigeon fed every fifteen seconds regardless of what it does is soon turning in circles between feedings. Phenomenon and researcher?

   Answer

   Superstitious behavior, Skinner (1948). Whatever the bird was doing when food arrived was accidentally reinforced, and the accident repeated.

8. Rats run a maze daily with no food and seem to wander. On day eleven food appears and their errors drop at once to the level of rats fed all along. Phenomenon and researchers?

   Answer

   Latent learning, Tolman and Honzik (1930). The rats had a cognitive map from the unrewarded days; reinforcement gave them a reason to show it.

9. Raccoons trained with food to drop coins in a container increasingly rub and dip the coins instead, delaying the food. Phenomenon and researchers?

   Answer

   Instinctive drift, Breland and Breland (1961). The trained response drifted toward innate food-washing at the cost of reinforcement: consequences shape behavior only within a species' evolved repertoire.

10. A dog wags and drools when it hears the treat pouch open, then sits on cue and gets the treat. Which parts are classical and which operant?

    Answer

    Both. Wagging and drooling at the pouch sound are a conditioned response to a stimulus that predicts food: classical. The sit is a voluntary behavior followed by a treat that depends on it: operant, positive reinforcement.

A study plan using this site

The teachers page gives a two-session route through the front page for a first pass. This is the exam-oriented version: quadrants and schedules first, a term check last. Spread it over a week.

1. The front page, first half. The one-paragraph version, the A-B-C, and the four quadrants with the interactive checker.
2. Negative reinforcement. The negative reinforcement page clears up the confusion the test writers use most.
3. Schedules, with the simulator. The schedules page; run all four until you can sketch each record. If one is shaky, its own page (FR, VR, FI, VI) has more examples.
4. Extinction and shaping. The extinction page for the burst, spontaneous recovery, and the partial reinforcement effect; the shaping page; then ten minutes in the lab shaping a lever press.
5. Operant vs. classical. The comparison page and its checklist. The page to reread the night before.
6. Names and dates. The history page, with the studies table above.
7. Practice. The examples page, the quiz, then the set above. For each wrong answer, find which of the four confusions produced it.
8. Term check. Go down the learning terms in the CED and give a one-line definition of each; the glossary has every one.^[1]

Common mistakes, and what you can skip

Common mistakes

• Classifying by how the consequence felt. A punisher is a consequence that reduced behavior; a reinforcer is one that increased it. Nothing else counts.
• Naming the schedule from the reinforcer. Money, praise, and food can be delivered on any schedule. Ask count or time, fixed or variable.
• Calling extinction forgetting, or punishment. Extinction is new learning that the response no longer pays, which is why it returns after a rest.
• Confusing a signal with a consequence. A stimulus before a reflex is a conditioned stimulus; a stimulus before a voluntary behavior that signals it will pay is a discriminative stimulus; a stimulus after the behavior is a consequence.
• Treating the contrast cases as refutations. Latent learning, insight, and observational learning show that learning also occurs without direct reinforcement, not that reinforcement fails to change behavior.

What the CED does not require

This site goes further than the AP course. Unless the current CED says otherwise, you do not need the following: the matching law; behavioral momentum; the response deprivation hypothesis behind the Premack principle; motivating operations; functional analysis; the equations of the Rescorla–Wagner model; autoshaping, sign-tracking, and contrafreeloading; and the debate over whether the positive/negative distinction should be kept at all.^[1] Reading them makes the required material easier to hold, and the neuroscience page connects reinforcement to the dopamine system covered elsewhere in the course, but none of it is where the points are.

Key takeaways

• In the revised AP Psychology course, learning is taught within Unit 3, Development and Learning, and the Course and Exam Description is the authority on which terms and question formats are tested.
• Every operant scenario is two questions — added or removed, and behavior up or down — and the answers name the quadrant; "positive" and "negative" describe the environment, not the outcome.
• Negative reinforcement is not punishment, and extinction is not a quadrant: punishment adds an aversive or removes a reinforcer after the response; in extinction the reinforcer simply stops arriving.
• Schedules are named by count-or-time and fixed-or-variable: pause-and-run on fixed ratio, a high steady rate on variable ratio, the scallop on fixed interval, a moderate steady rate on variable interval.
• Learn the classic studies as procedures with findings, because the exam describes what was done and asks what it showed; latent learning, insight, and the Bobo doll are the contrast cases, and instinctive drift and preparedness mark the biological limits.

Check yourself

A teacher says she "negatively reinforced" a student by assigning detention, and the student's talking in class decreased. What did she actually do?

Positive punishment. Something was added (detention) and the behavior went down, so the consequence was a punisher, and because it was added rather than removed it was positive. Negative reinforcement would require the behavior to increase because something was taken away.

One rat gets a pellet after every twentieth press; another gets a pellet for the first press after twenty seconds have passed. Sketch the cumulative record you expect from each.

The first rat is on a fixed ratio 20 and should show break-and-run: a pause after each pellet, then a steep run of twenty presses. The second is on a fixed interval 20 seconds and should show a scallop: few presses just after each pellet, then an accelerating rate as the twenty seconds run out. The ratio schedule produces the higher overall rate, because responding faster brings the pellet sooner; on the interval schedule it does not.

Children who watched an adult hit a Bobo doll later hit it themselves. A classmate calls this operant conditioning because "aggression was reinforced." Is that right?

No. In the 1961 study the children were not reinforced for imitating; they reproduced the specific acts they had watched. That is observational learning, and it is on the exam precisely as the contrast with learning through one's own consequences. Operant conditioning would require the children's own hitting to have been followed by a consequence that changed its frequency.

Want more? The 20-question quiz covers every page on the site with instant explanations.