Open your phone and pull down to refresh a feed that's already up to date. Why do you do that — repeatedly, almost without deciding to?
A pigeon in a Harvard basement in the 1950s would understand completely. B.F. Skinner put hungry pigeons in a box where pecking a disk sometimes delivered food — not every time, and never on a schedule you could predict. The pigeons didn't peck calmly. They pecked fast, steadily, and for a maddeningly long time after the food stopped coming entirely. Skinner had built, in a box, the exact machinery now living in your pocket: a behavior whose reward arrives unpredictably, which turns out to be the single most addictive reward arrangement psychology has ever found.
That's the engine of this lesson. Classical conditioning (last lesson) is about reflexes you can't control — salivation, fear. Operant conditioning is about the voluntary behaviors you choose to repeat or abandon based on their consequences. Your brain is running a constant cost-benefit ledger on its own actions, and Skinner figured out the accounting rules.
Before Skinner, there was Edward Thorndike (1898), who put cats in puzzle boxes — crates a cat could escape only by pulling a loop or stepping on a lever, with a fish dinner waiting outside. The first escape was pure accident. But over repeated trials, the cat's flailing got shorter and the successful move came faster, until the cat would trip the latch almost immediately.
From this Thorndike drew the law of effect: behaviors followed by satisfying consequences become more likely to recur, and behaviors followed by unpleasant consequences become less likely. That one sentence is the seed of everything below. Notice what it is not: the cat isn't reasoning out the mechanism. The consequence reaches back and stamps in (or stamps out) the behavior automatically.
Skinner industrialized Thorndike's idea. He built the operant chamber (the "Skinner box") — a controlled environment with a lever or disk, a food/water dispenser, and often a way to deliver mild shock, plus instruments that automatically recorded every response. This let him measure behavior precisely and manipulate consequences cleanly.
Skinner's master term: operant behavior is behavior that "operates" on the environment to produce consequences. And a reinforcer is any consequence that increases the behavior it follows; a punisher is any consequence that decreases the behavior it follows. Read those definitions again, because they're functional — defined by their effect, not by whether they sound pleasant. Something only counts as a reinforcer if behavior actually goes up.
This is the single most tested, most misunderstood diagram in AP Psychology. Two questions, asked independently:
Here, "positive" and "negative" do not mean good and bad. They mean plus (add something) and minus (take something away) — like math signs. Burn that in now and you'll dodge the most common error on the exam.
| **Add a stimulus (positive)** | **Remove a stimulus (negative)** | |
|---|---|---|
| Behavior INCREASES (reinforcement) | Positive reinforcement: add something desirable. Get a treat for cleaning your room → you clean more. | Negative reinforcement: remove something aversive. Buckle up to stop the annoying beep → you buckle more. |
| Behavior DECREASES (punishment) | Positive punishment: add something aversive. Get a speeding ticket → you speed less. | Negative punishment: remove something desirable. Lose phone privileges → you break curfew less. |
The cell students butcher is negative reinforcement. It is not punishment. You remove something unpleasant (the seatbelt beep, the headache when you take aspirin, the nagging when you finally do the dishes) and the behavior that turned it off gets stronger. Negative reinforcement is the psychology of relief and escape. Aspirin works by negative reinforcement: taking it removes pain, so you take it again next headache.
Try This. For each, name the quadrant: (1) Your dog jumps up, you turn and ignore it, the jumping fades. (2) You put on sunscreen and the sunburn never comes, so you keep wearing it. (3) A toddler throws food and gets sent to time-out, ending the tantrums. (4) You text a friend and get a funny reply, so you text more.
Answers: (1) negative punishment — removing attention, jumping decreases; (2) negative reinforcement — avoiding/removing an aversive sunburn, behavior increases; (3) negative punishment — removing freedom/fun, tantrums decrease; (4) positive reinforcement — adding a reward, texting increases.
A primary reinforcer is innately satisfying — it works without any learning because it meets a biological need: food, water, warmth, relief from pain. A secondary (or conditioned) reinforcer gains its power by association with primary reinforcers: money, grades, gold stars, the click of a clicker in dog training. Money buys nothing on a desert island; it's reinforcing only because it's been paired with everything it can purchase.
Skinner himself was skeptical of punishment, and the exam wants you to know why. Punishment suppresses a behavior without teaching a replacement (it says "stop," never "do this instead"). It can trigger fear and aggression that generalize to the punisher. It often works only while the punisher is watching, breeding sneakiness rather than change. And severe punishment can model aggression as a problem-solving tool. Reinforcing a desirable alternative is usually more effective than punishing the undesirable one.
How do you reinforce a rat for pressing a lever it has never once pressed? You don't wait forever. You use shaping: reinforcing successive approximations — successively closer steps toward the target behavior. First you reinforce the rat for facing the lever, then for approaching it, then for touching it, then for pressing it. Each reward raises the bar. This is how animal trainers get dolphins to flip and how teachers build complex skills one piece at a time.
Once a behavior exists, how often you reinforce it dramatically shapes how it's performed. Continuous reinforcement rewards every correct response — fastest way to learn a new behavior, but it also extinguishes fast (the moment rewards stop, the animal quits, because the change is obvious). Partial (intermittent) reinforcement rewards only some responses — slower learning, but far more resistant to extinction.
Partial schedules split two ways — by ratio (number of responses) vs. interval (passage of time), and by fixed (predictable) vs. variable (unpredictable):
Mnemonic: ratio → rapid (paid per response, so respond more); variable → vigorous and persistent (unpredictability keeps you going). Your refresh-the-feed habit from the hook is variable-ratio — which is exactly why it's so hard to stop.
In operant conditioning, extinction is the weakening and disappearance of a behavior when reinforcement stops. You may first see an extinction burst — a brief spike in the behavior (the vending machine ate your dollar, so you hit the button harder first). Variable schedules drag extinction out longest because the absence of reward looks, for a while, just like a normal dry spell.
The Premack principle says a more-preferred activity can reinforce a less-preferred one: "Eat your vegetables (low-frequency) and then you may have dessert (high-frequency)." Grandma's Rule. The behavior you like becomes the reinforcer for the behavior you don't.
Operant principles run real systems. A token economy gives secondary reinforcers (tokens, points, stars) for desired behaviors, later exchangeable for primary reinforcers or privileges — used in classrooms, psychiatric wards, and rehab. Behavior modification more broadly applies reinforcement and shaping to change behavior systematically. The same logic governs loyalty cards, video-game loot boxes, and your fitness app's streak counter.
Skinner's operant chamber and the schedules of reinforcement (B.F. Skinner, 1930s–1950s).
Who & when: B.F. Skinner, working at Minnesota and then Harvard from the 1930s onward, building on Thorndike's law of effect.
Method: Skinner placed a rat or pigeon in an operant chamber equipped with a lever (rats) or pecking disk (pigeons), a food dispenser, and an automatic recorder that drew a cumulative record — a line that steps upward with each response, so a steep line means fast responding and a flat line means none. By varying when and how often food followed responses, Skinner systematically compared reinforcement schedules.
What he found: The schedule, not just the reward, controlled behavior. Variable-ratio schedules produced the highest, most persistent response rates; fixed-interval schedules produced the telltale scalloped curve — a pause after reinforcement, then a rush as the next interval elapsed. Partially reinforced behaviors resisted extinction far longer than continuously reinforced ones (the partial reinforcement effect).
Why it matters: Skinner converted Thorndike's loose "law of effect" into a precise, predictive science of consequences. The cumulative-record curves for the four schedules are a fixture of the AP exam, and the principles explain everything from gambling addiction to classroom token economies. For the exam: Skinner = operant chamber, reinforcement schedules, shaping, and "behavior is shaped by its consequences."
Scenario 1. Maria's little brother whines in the grocery store until she buys him candy. The next week, he whines louder and longer. Meanwhile, Maria notices that buying the candy makes the whining stop instantly, so she finds herself buying candy faster each trip.
Name the process for each person. Two operant processes are running at once. For the brother, whining is strengthened by positive reinforcement — he whines and gets candy added, so whining increases. For Maria, buying candy is strengthened by negative reinforcement — buying removes the aversive whining, so her candy-buying increases. This "reinforcement trap" is why the cycle escalates: each person's behavior reinforces the other's.
Scenario 2. A teacher gives students a sticker for every book they finish. Soon the class reads voraciously. Midyear, the teacher switches to a system where students earn a sticker only after an unpredictable number of books — sometimes two, sometimes six. Reading stays high, and when the teacher forgets stickers for a week, students keep reading anyway.
Identify the schedules and explain the persistence. The teacher started on continuous reinforcement (a sticker for every book — great for establishing the behavior). The switch to "an unpredictable number of books" is a variable-ratio schedule. Because reinforcement is now unpredictable, the behavior becomes highly resistant to extinction — students keep reading through the stickerless week because, on a VR schedule, a dry spell is indistinguishable from normal.
Scenario 3. A dog trainer wants a dog to roll over — a behavior the dog has never performed. She gives a treat first for lying down, then only for lying on its side, then only for a partial roll, then only for a full roll.
What technique is this, and what are the intermediate rewarded behaviors called? This is shaping. The intermediate behaviors (lie down → lie on side → partial roll) are successive approximations — each a closer step to the target, reinforced in turn until the full behavior emerges.
Negative reinforcement vs. punishment — the #1 AP confusion. "Negative" tempts students to call it punishment. But negative reinforcement always increases behavior by removing something aversive (take aspirin → headache gone → take aspirin again). Punishment decreases behavior. Test tip: find the behavior, ask "did it go up or down?" Up = reinforcement, every time — no matter how unpleasant the situation sounds.
Positive punishment vs. negative reinforcement. Both involve an aversive stimulus, so they blur. Positive punishment adds the aversive thing to stop a behavior (spanking, a ticket → behavior down). Negative reinforcement removes an already-present aversive thing to strengthen a behavior (seatbelt beep stops → buckling up). Add-to-stop vs. remove-to-strengthen.
The four schedules. Mix up ratio and interval and you'll mis-predict the curve. Ratio = responses (rate depends on you); interval = time (rate depends on the clock). Variable = unpredictable = persistent. Memorize that variable-ratio is the highest rate + most extinction-resistant, and fixed-interval makes the scalloped curve.
Reinforcement vs. reward. A reward is a reinforcer only if it actually increases the behavior. If you "reward" a kid with broccoli he hates and his behavior doesn't change, it wasn't a reinforcer. Reinforcers are defined by their effect, not by whether they sound nice.
Four-choice MCQs in current AP format. Answers and explanations in section (h).
` Cumulative | ____ Curve W (steady & steep) responses | ____/ | ___/ ___ Curve X (scalloped) | __/ __/‾‾ | __/‾‾ __/‾‾ | _/‾‾ __/‾‾ |_/__/‾‾_________________________ Time ` Curve W (steady and steep) and Curve X (scalloped) most likely represent which schedules, respectively?EBQ format: three summarized peer-reviewed sources on a shared topic. Write a response that (1) states a defensible claim, (2) supports it with evidence from at least two of the sources, and (3) provides reasoning that explains how the evidence supports the claim using course content. 7 points total: Claim (0–1) + Evidence from 2+ sources (0–3) + Reasoning & Application (0–2).
Prompt. Using the three sources below, develop an argument about whether reinforcement or punishment is the more effective tool for producing lasting behavior change.
Source 1 — Skinner (behavioral analysis).
A review of operant-conditioning research argues that punishment tends to suppress unwanted behavior only temporarily and only in the presence of the punishing agent. The review reports that animals and children whose unwanted behaviors were punished frequently resumed those behaviors once the punisher was absent, and often showed increased fear and aggression directed at the punisher. By contrast, behaviors that were reinforced — especially desirable alternatives to the unwanted behavior — were maintained more durably and generalized to new settings.
Source 2 — token-economy field study.
Researchers ran a token economy in three middle-school classrooms. Students earned tokens (exchangeable for privileges) for on-task behavior; a comparison set of classrooms used a demerit system that removed recess minutes for off-task behavior. On-task behavior rose 41% in the token (reinforcement) classrooms versus 12% in the demerit (punishment) classrooms. However, the demerit classrooms did show a rapid short-term drop in disruptions during the first two weeks before the effect plateaued.
Source 3 — parental-discipline longitudinal study.
A longitudinal study followed 600 families over four years. Families relying primarily on physical/positive punishment reported faster immediate compliance but higher rates of child aggression and parent-child conflict by year four. Families relying primarily on reinforcement of prosocial behavior reported slower initial compliance but more stable long-term behavior and lower aggression. The authors note that some punishment-reliant families also used inconsistent reinforcement, complicating the comparison.
A. Propose a specific, defensible claim. B. Support your claim with evidence from at least two sources. C. Explain how that evidence supports your claim, using at least one operant-conditioning concept.
Claim. Reinforcement is the more effective tool for producing lasting behavior change, whereas punishment tends to produce only short-term suppression. (Claim: 1 pt — defensible and directly answers the prompt.)
Evidence. Source 2 found that on-task behavior rose 41% in the reinforcement (token-economy) classrooms compared with only 12% in the punishment (demerit) classrooms, indicating reinforcement produced the larger behavioral gain. Source 3 reinforces this over a longer span: families relying on reinforcement of prosocial behavior reported more stable long-term behavior and lower child aggression by year four, while punishment-reliant families showed faster immediate compliance but higher aggression and conflict over time. (Evidence: 3 pts — accurate, specific evidence drawn from two different sources, each clearly tied to the claim, including the cited figures.)
Reasoning & Application. These findings fit the operant principle that a reinforcer strengthens and maintains the behavior it follows, while punishment merely suppresses a behavior without teaching a replacement. In the token economy, the tokens act as secondary (conditioned) reinforcers, building durable on-task behavior; the demerit system only removes a privilege to suppress disruption, so it fades once the threat is gone — consistent with Skinner's argument in Source 1 that punished behavior resumes in the punisher's absence and can breed fear and aggression. Because reinforcement establishes a behavior that is actively maintained (rather than one that is only suppressed under threat), it generalizes better and lasts longer — exactly the long-term stability Source 3 documents. (Reasoning & Application: 2 pts — explains the causal link AND correctly applies operant concepts: reinforcement maintains vs. punishment suppresses, plus secondary reinforcers.)
1. (D) Negative punishment. A desirable stimulus (tablet privileges) is removed and the behavior (tantrums) decreases → negative (removal) punishment (decrease). (A)/(B) are reinforcement, which would increase behavior. (C) positive punishment would add an aversive stimulus, not remove a desirable one.
2. (B). The law of effect: satisfying consequences make a behavior more likely; unpleasant ones less likely. (A) is observational learning (Bandura, L20); (C) is classical conditioning; (D) describes variable schedules, not Thorndike's law.
3. (B) Negative reinforcement. The antacid removes an aversive stimulus (heartburn) and the behavior (taking antacids) increases → negative reinforcement. This is the classic relief/escape case. Not punishment — the behavior went up, not down.
4. (B) Variable-ratio. VR produces the highest, steadiest response rate and the most resistance to extinction because reinforcement is unpredictable (the slot-machine effect). The others all yield lower or less persistent responding.
5. (C) Fixed-ratio. Reinforcement after a set number of responses (every 20 units) is fixed-ratio, which yields a high rate with a short post-reinforcement pause. (A)/(B) are interval (time-based) schedules; (D) misnames it — the schedule is fixed, not variable.
6. (B) Shaping. Reinforcing successive approximations toward a target behavior is shaping. (A) extinction weakens behavior; (C) spontaneous recovery is a classical-conditioning term; (D) generalization is responding to similar stimuli.
7. (A) Premack principle. Using a high-frequency, preferred activity (video games) to reinforce a low-frequency one (homework) is the Premack principle. (B) negative reinforcement removes an aversive stimulus; (C) a token economy uses exchangeable tokens; (D) concerns extinction resistance.
8. (B) Secondary (conditioned) reinforcers. Money, grades, and chips have no innate value; they're reinforcing through learned association with primary reinforcers. (A) primary reinforcers satisfy biological needs directly; (C)/(D) are the wrong category entirely.
9. (B) Token economy. Earning tokens for desired behavior and exchanging them for reinforcers is the textbook token economy. (A) is a schedule, not a system; (C) negative punishment would remove something to decrease behavior; (D) latent learning is unrelated (L21).
10. (B) Variable-ratio and fixed-interval. The steady, steep straight line (Curve W) is the signature of variable-ratio — high, constant responding with no pauses. The scalloped curve (Curve X) — pause after each reinforcement, then acceleration toward the deadline — is the signature of fixed-interval. (A) reverses them; (C) and (D) misassign at least one curve.
11. (B) Positive punisher. Extra chores are added (positive) and the behavior (missing curfew) decreases → positive punishment. (A) negative reinforcement would remove something and increase behavior; (C)/(D) are reinforcer types, but the behavior decreased, so it isn't reinforcement at all.
12. (C). Skinner emphasized that punishment suppresses behavior without teaching a replacement and can provoke fear/aggression. (A) is false — punishment doesn't erase memory; (B) misstates the issue (the problem is no replacement is taught at all); (D) is fabricated.
13. (D) Variable-interval. Reinforcement for the first response after an unpredictable amount of time yields a slow, steady rate — exactly the email-checking pattern. (C) fixed-interval would produce scalloping; (A)/(B) are ratio (response-count) schedules, but here the payoff depends on time, not number of checks.
14. (B) Extinction burst. A brief increase in a behavior right after reinforcement stops, before it fades, is an extinction burst. (A) spontaneous recovery is the reappearance of an extinguished behavior after a rest; (C) and (D) are unrelated processes.
15. (B) Partial reinforcement effect. Behaviors reinforced on partial (here variable-ratio) schedules resist extinction longer than continuously reinforced ones — the partial reinforcement effect. (A) the law of effect is the general principle, not this specific contrast; (C)/(D) don't address extinction resistance.
| Component | Points | Criteria |
|---|---|---|
| Claim | 0–1 | 1 pt for a defensible claim that directly answers whether reinforcement or punishment is more effective. No point for restating the prompt or hedging both ways. |
| Evidence | 0–3 | Up to 3 pts for accurate, specific evidence drawn from at least two different sources and connected to the claim. Evidence from only one source caps this at a lower score; cited figures strengthen it. |
| Reasoning & Application | 0–2 | Up to 2 pts for explaining how the evidence supports the claim using correct operant-conditioning course content (e.g., reinforcement maintains vs. punishment suppresses; secondary reinforcers; drawbacks of punishment). No credit for unexplained assertion. |
PsyIQ · Lesson 19 of 30 · Unit 3: Development and Learning. FRQ practice is an Evidence-Based Question (EBQ) modeled on the redesigned (2025+) AP Psychology exam. Not affiliated with the College Board. AP is a registered trademark of the College Board. Content pending external psychology QC.