AP Psychology · Lesson 20 of 30
PsyIQ · AP Psychology

Lesson 20: Observational & Cognitive Learning

Unit 3 · Development and Learning (15–25%) · Science Practices:** 1 — Concept Application (primary); 4 — Argumentation (FRQ); 3 — Data Interpretation (supporting)
Objectives:
  • Explain how organisms learn by *watching* others — Bandura's observational learning, vicarious reinforcement, and the four processes (attention, retention, reproduction, motivation) — without any direct reinforcement of their own.
  • Distinguish cognitive forms of learning (Tolman's latent learning and cognitive maps, Köhler's insight) from simple stimulus-response conditioning, and explain what the "cognitive" part adds.
  • Apply intrinsic vs. extrinsic motivation, the overjustification effect, and Seligman's learned helplessness to real scenarios, and tell each apart from the things students confuse them with.

(a) Hook

You have never punched a stranger, set off a fire alarm for fun, or tried to land a kickflip — and yet you know, in vivid detail, how each one would go. Where did that knowledge come from? Not from being rewarded for doing them. You learned by watching.

This is the part of learning that pure behaviorism couldn't explain. Skinner's rats had to actually press the lever and get the pellet. But humans (and plenty of other animals) pick up entire behaviors having never performed them once — never reinforced, never punished, sometimes never even consciously trying to learn. A toddler swears the first time after hearing it one time in traffic. A teenager mimics a YouTuber's slang within a week.

The psychologist who forced this into the field was Albert Bandura, and his weapon was an inflatable clown. In this lesson, learning stops being something that happens to you through reward and punishment and becomes something your mind does — watching models, building maps, having flashes of insight, and deciding whether the payoff is even worth it.

(b) Core Concepts

Learning by watching: observational learning

Observational learning (also called social learning) is learning by observing and imitating others — acquiring new behaviors without direct experience of reinforcement or punishment. The person being watched is the model. You don't have to do the behavior, and you don't have to be rewarded for it; watching is enough to acquire it.

The crucial mechanism is vicarious reinforcement: when you see a model get rewarded for a behavior, you become more likely to imitate it — as if you'd been reinforced yourself. The flip side is vicarious punishment: seeing a model get punished makes you less likely to copy the behavior. Notice the word vicarious — the consequence happens to someone else, but it changes your behavior. This is how you learn that the express checkout line gets you yelled at if you have 20 items: you watched it happen to the person ahead of you.

A subtle but exam-critical point: observational learning explains how a behavior gets learned (acquired), which is different from whether it gets performed. Bandura showed that children learned aggressive acts just by watching, even when they didn't immediately copy them — the knowledge was banked, ready to come out later if the situation rewarded it.

Bandura's four processes

Bandura argued that for observational learning to actually produce a behavior, four things must happen. Memorize them as A-R-R-M (Attention, Retention, Reproduction, Motivation):

Try This. Think of a skill you learned almost entirely by watching — tying a tie, a dance move, a video-game combo. Walk it through A-R-R-M. Which model did you attend to (and why them)? How did you retain the steps? Where did reproduction break down (the part your hands couldn't do yet)? What motivated you to actually try it? If one stage failed, that's exactly where your learning stalled.

Mirror neurons (handle with care)

You may hear that observational learning is powered by mirror neurons — neurons (first reported in monkeys' premotor cortex by Rizzolatti's lab in the 1990s) that fire both when an animal performs an action and when it watches another perform the same action. The popular story says they're the brain's imitation-and-empathy machine. Be cautious: while mirror neurons are real in monkeys, the claim that they explain human imitation, empathy, and culture is scientifically debated and often overstated. For the exam, know the term and the basic definition, but don't treat mirror neurons as the proven cause of observational learning.

Modeling that shapes society: prosocial and antisocial effects

Because models teach so efficiently, what gets modeled matters enormously. Prosocial modeling — demonstrating helpful, cooperative, generous behavior — reliably increases those behaviors in observers; this is the logic behind mentorship, "be the change," and TV programs designed to teach sharing. Antisocial modeling does the reverse: observing aggression, cruelty, or rule-breaking can teach and disinhibit those behaviors.

This is the heart of the media effects debate. Research generally finds that watching violent media is associated with increased aggression in viewers (a real, replicated correlation, and supported by experiments like Bandura's). But "associated with" is doing careful work — media is one influence among many, effect sizes are modest, and violent media does not turn ordinary people into violent criminals. The honest exam answer: media models can increase aggression, especially when violence is rewarded or goes unpunished onscreen, but it is one contributing factor, not a sole cause.

Cognitive learning: the mind builds maps

Strict behaviorists treated the mind as a black box. Cognitive learning pried it open, showing that learning involves mental processes — expectations, representations, understanding — not just stamped-in stimulus-response links.

The pioneer was Edward Tolman. Running rats through mazes, Tolman discovered latent learning — learning that occurs without reinforcement and stays hidden ("latent") until there's a reason to demonstrate it. Rats allowed to wander a maze with no food reward seemed to learn nothing — until food was suddenly introduced, at which point they raced to the goal almost immediately, as fast as rats that had been rewarded all along. They'd been learning the maze the whole time; the reward just gave them a reason to show it. (Notice the deep parallel to Bandura's learning/performance distinction — reinforcement affects performance, not necessarily acquisition.)

Tolman argued the rats had built a cognitive map — a mental representation of the spatial layout of their environment. They weren't blindly chaining left-right turns; they knew where things were. Underlying all of this is expectancy: the rats developed expectations about what leads to what, and learning meant updating those expectations. This is why Tolman's approach is sometimes called purposive behaviorism — behavior aimed at goals, guided by cognitive expectations.

Insight: the sudden click

Wolfgang Köhler studied problem-solving in chimpanzees and described insight learning — solving a problem through a sudden flash of understanding rather than gradual trial-and-error. His chimp Sultan, unable to reach a banana hung out of reach, paused, then suddenly stacked boxes or fitted two sticks together to grab it — as if the solution arrived all at once. Insight is the cognitive opposite of slow operant shaping: no incremental reinforcement of approximations, just a reorganized understanding of the problem. That "aha!" you feel when a puzzle suddenly cracks is insight.

Motivation: intrinsic, extrinsic, and the reward that backfires

Why we do things shapes how learning sticks. Intrinsic motivation is the desire to perform a behavior for its own sake — because it's interesting, satisfying, or enjoyable. Extrinsic motivation is the desire to perform a behavior to gain an external reward or avoid punishment — money, grades, praise, trophies.

Here's the counterintuitive part the exam loves: the overjustification effect occurs when offering an external reward for an activity someone already enjoys reduces their intrinsic motivation. Pay kids to draw pictures they already loved drawing, and once you stop paying, they draw less than kids who were never paid. The external reward "overjustifies" the behavior — the person re-interprets "I do this because I love it" as "I do this for the reward," so when the reward vanishes, so does the drive. Extrinsic rewards aren't always harmful (they're great for boring or new tasks), but bolting them onto an already-loved activity can quietly kill the love.

Learned helplessness

Finally, Martin Seligman discovered learned helplessness — the passive resignation an organism learns when it has been exposed to repeated, unavoidable aversive events, eventually failing to act even when escape later becomes possible. Dogs given inescapable shocks later didn't even try to jump a low barrier to escape shocks they easily could have avoided; they'd learned that their actions didn't matter. Seligman connected this to human depression: when people repeatedly experience outcomes they feel powerless to control, they may stop trying altogether. The key ingredient is perceived lack of control — not lack of ability.

(c) Classic Studies Spotlight

Bandura's Bobo Doll Experiment (Bandura, Ross & Ross, 1961).

Who & when: Albert Bandura and colleagues, Stanford University, 1961.

Method: Nursery-school children (ages roughly 3–6) were assigned to conditions. In the key conditions, a child watched an adult model interact with a five-foot inflatable "Bobo" doll. In the aggressive condition, the adult attacked the doll — punching it, kicking it, hitting it with a mallet, and shouting distinctive phrases like "Sock him in the nose!" In the nonaggressive condition, the model played quietly and ignored the doll; a control group saw no model. Each child was then mildly frustrated (shown attractive toys, then told those were for other children) and left alone in a room that contained a Bobo doll and other toys. Observers recorded the child's behavior through a one-way mirror.

Results: Children who had watched the aggressive model imitated the model's specific physical and verbal attacks far more than children in the other groups — often reproducing the exact actions and phrases. Boys showed more physical aggression than girls overall, and aggression was higher when the model was the same sex as the child. In a later variation, children who saw the model punished for aggression imitated less than those who saw the model rewarded or face no consequences — but when later offered an incentive to reproduce the behavior, all groups could do so, proving they had learned it regardless of the consequence they'd witnessed.

Why it matters: Bobo demonstrated that aggression — and behavior generally — can be learned through observation alone, with no direct reinforcement of the learner, and showcased vicarious reinforcement/punishment and the learning-vs-performance distinction. It cracked open behaviorism and launched social learning theory.

(d) Application Practice

Scenario 1. Maya watches her older sister get extra dessert every time she clears her plate without complaining. Within a week, Maya — who was never herself rewarded — starts clearing her plate cheerfully too.

Which concept, and what's the tell? This is observational learning driven by vicarious reinforcement. The tell: Maya changes her behavior after seeing someone else rewarded, never being reinforced herself. (Trace A-R-R-M if you like: she attended to a high-status similar model, retained the behavior, could reproduce it, and the witnessed reward supplied motivation.)

Scenario 2. A delivery driver spends months driving all over a city for work, never trying to memorize the layout. One day his GPS dies, and to his own surprise he navigates to a brand-new address quickly and accurately, picturing the streets in his head.

Which concept? Latent learning, revealed by a cognitive map. He acquired spatial knowledge with no reinforcement for learning it; the dead GPS (a new demand) is the moment that latent learning finally gets performed. He's the human version of Tolman's rat.

Scenario 3. Devon has loved playing piano since he was little. His parents start paying him $5 per practice session to "encourage" him. Months later they stop the payments — and Devon now practices noticeably less than before they ever paid him.

Which concept? The overjustification effect. An extrinsic reward (money) was attached to an already intrinsically motivating activity (piano he loved), undermining the intrinsic motivation; when the reward stopped, the original drive had eroded.

Scenario 4. After a year of a chaotic job where nothing she did improved her impossible workload, Priya transfers to a new team with reasonable goals — yet she still doesn't speak up or attempt projects she's clearly capable of, assuming nothing will work anyway.

Which concept? Learned helplessness (Seligman). Repeated exposure to uncontrollable aversive conditions taught passivity that persists even after control returns. The barrier is perceived powerlessness, not lack of skill — she's fully capable.

(e) Traps & Confusions

Observational learning vs. operant conditioning. Both change behavior via consequences, but in operant conditioning you perform the behavior and you receive the consequence (Skinner's lever-and-pellet). In observational learning, someone else performs it and receives the consequence, and you learn by watching (vicarious reinforcement). Mnemonic: operant = "I did it and got the result"; observational = "I watched them get the result."

Latent learning vs. reinforced learning. The trap is thinking no reward means no learning. Latent learning proves learning happens without reinforcement — it's just hidden until there's a reason to show it. Reinforcement affects performance, not whether the learning occurred. Tolman's unrewarded rats had learned the maze all along.

Intrinsic vs. extrinsic / the overjustification trap. Intrinsic = for its own sake (you'd do it free); extrinsic = for an outside reward. The overjustification effect is the specific case where an extrinsic reward kills intrinsic motivation for an already-enjoyed activity. Don't apply it to boring tasks no one liked to begin with — there's no intrinsic motivation there to undermine.

Learned helplessness vs. lack of skill. Learned helplessness is learned passivity from perceived lack of control — the person could succeed but has stopped trying. If someone fails because they genuinely can't do the task (no ability), that's not learned helplessness. The signature of helplessness is giving up when success is actually possible.

(f) Practice Problems

Four-choice MCQs in current AP format. Answers and explanations in section (h).

Question 1
A child learns to fear the stove after watching her brother get burned and cry, even though she was never burned herself. This best illustrates
Question 2
According to Bandura, which is the correct order of the four processes required for observational learning to produce behavior?
Question 3
In the Bobo doll variation where children saw the model punished, the children imitated less aggression — but when later offered a reward to reproduce the behavior, they all could. This best demonstrates the distinction between
Question 4
Tolman's rats that explored a maze with no reward, then ran it efficiently once food appeared, demonstrated
Question 5
A student who genuinely loves writing poetry begins receiving cash prizes for it. After the contest ends, she writes far less than before. This is best explained by
Question 6
Köhler's chimpanzee suddenly stacking boxes to reach a banana, after a pause and with no gradual trial-and-error, is the classic example of
Question 7
A dog repeatedly given shocks it cannot escape later fails to jump an easily cleared barrier to avoid shock. Seligman called this
Question 8
Which statement about mirror neurons reflects current scientific caution?
Question 9
A new employee learns the unwritten rule "don't interrupt the boss in meetings" by watching a coworker get coldly shut down for doing so. The employee's behavior change is driven by
Question 10
Data interpretation. In a study modeled on Bobo, mean imitative-aggression acts were recorded by condition: Aggressive-model = 12.4, Nonaggressive-model = 1.8, No-model control = 2.0. The most defensible conclusion is that
Question 11
Which scenario best illustrates a cognitive map rather than simple stimulus-response learning?
Question 12
A coach wants players to keep enjoying a drill they already love. Based on the overjustification effect, she should be most cautious about
Question 13
Data interpretation. Two rat groups run a maze for 17 days. Group A is rewarded every day; its error count drops steadily from the start. Group B gets no reward until Day 11; its error curve stays high and flat through Day 10, then plunges sharply on Day 12 to match Group A. The Day-11-to-12 plunge is best explained by
Question 14
A teacher notices a capable student has stopped attempting any problems after a semester in a previous class where, no matter how hard she worked, she was told her answers were wrong. The student's behavior most likely reflects
Question 15
Which of the following is the clearest difference between observational learning and operant conditioning?

(g) FRQ Practice — Article Analysis Question (AAQ)

Respond to all six parts (A–F) in complete sentences using appropriate psychological terminology.

Stimulus — summarized study

Introduction. Researchers examined whether watching a peer model be rewarded versus punished for a behavior affects how often young children imitate that behavior, and whether children have nonetheless learned the behavior regardless of the consequence they observed.

Participants. 120 children (ages 4–5; 50% girls, 50% boys) were recruited from four preschools in a single mid-sized U.S. city. Reported race/ethnicity was 60% White, 18% Hispanic/Latino, 12% Black, 7% Asian, and 3% Other. Parents gave written informed consent, and each child was identified in the dataset only by a randomly assigned code number.

Method. Each child watched a short video of a same-age peer playing a new toy by performing a distinctive novel action (twisting a dial three times before pressing a button). Children were randomly assigned to one of two conditions: in the Reward condition the video ended with an adult praising the peer and giving a sticker; in the Punishment condition the video ended with the adult scolding the peer and taking the toy away. Each child was then left alone with the toy for three minutes (a "free-play" measure), and observers behind a one-way mirror counted how many times the child spontaneously performed the distinctive action. Afterward, an experimenter offered every child a sticker to "show me how the kid in the video played with the toy" (a "prompted" measure), and again counted the distinctive actions.

Results. During free play, children in the Reward condition performed the distinctive action a mean of 4.6 times, versus 1.3 times in the Punishment condition. On the prompted measure, the two groups were nearly identical: 5.1 (Reward) versus 4.9 (Punishment).

A. Identify the research method used in this study.

B. State the operational definition of the dependent variable measured during free play.

C. Describe what the free-play means indicate about the difference between the Reward and Punishment conditions. (Cite the numbers.)

D. Identify one ethical guideline the researchers applied, and describe how they applied it.

E. Explain the extent to which the findings are or are not generalizable, using specific evidence from the study.

F. Explain how the pattern across the free-play and prompted measures supports or refutes a claim about observational learning. Apply a relevant psychological concept. (2 points)

Model answer (earns 7/7)

A. The research method was an experiment, because children were randomly assigned to conditions and the researchers manipulated an independent variable (whether the peer model was rewarded or punished). (1 pt — names the method)

B. The free-play dependent variable was operationally defined as the number of times a child spontaneously performed the distinctive action (twisting the dial three times then pressing the button) during the three-minute period alone with the toy, as counted by observers behind a one-way mirror. (1 pt)

C. The means indicate that children who saw the peer rewarded imitated the action substantially more during free play than children who saw the peer punished — a mean of 4.6 versus 1.3, a difference of about 3.3 actions — suggesting that the consequence the model received influenced how often children chose to perform the behavior. (1 pt — cites the numbers and direction)

D. The researchers applied informed consent: they obtained written consent from each child's parent before the child participated. (Confidentiality would also earn the point — each child appeared in the dataset only by a randomly assigned code number, with no names.) (1 pt)

E. The findings have limited generalizability because all participants were 4-to-5-year-old children drawn from only four preschools in a single U.S. city, so the results may not extend to older children, adults, or children in other cultures or regions; however, the sample was balanced by sex and included a range of racial/ethnic backgrounds, which supports somewhat broader applicability within that age group. (1 pt — commits to a direction, backed by study evidence)

F. The pattern supports the claim that observational learning produces learning (acquisition) that is separable from performance. Although the Reward group out-imitated the Punishment group during free play (4.6 vs. 1.3), once every child was offered a sticker to demonstrate the action, both groups performed it at nearly identical, high rates (5.1 vs. 4.9). This shows that children in the Punishment condition had in fact learned the behavior through observation just as well — they simply lacked the motivation to perform it until an incentive was provided. This illustrates vicarious punishment suppressing performance (the witnessed consequence to the model reduced spontaneous imitation) while leaving acquisition intact, exactly the learning-versus-performance distinction Bandura demonstrated with the Bobo doll. (2 pts — states support AND applies genuine psychological concepts: learning/performance distinction and vicarious punishment/motivation)

Where students commonly lose points

🔑 Answer Key

1. (B) Vicarious punishment in observational learning. She learns to fear the stove by seeing someone else suffer the consequence. (A) requires her own behavior-consequence link; (C) and (D) are unrelated to observed aversive consequences.

2. (B) Attention → Retention → Reproduction → Motivation. Bandura's four processes in order (A-R-R-M). The other sequences scramble them.

3. (C) Learning (acquisition) and performance. All children had learned the behavior (they could perform it for a reward), but witnessed punishment suppressed spontaneous performance. (A), (B), and (D) don't capture "learned it but didn't show it."

4. (B) Latent learning. Learning occurred without reinforcement and stayed hidden until food gave a reason to perform. (A) is operant shaping; (C) is a classical-conditioning term; (D) is sudden insight, not gradual unrewarded learning.

5. (B) The overjustification effect. An extrinsic reward (cash) attached to an already-loved activity undermined her intrinsic motivation, so output fell when the reward stopped. (A) names the starting motivation but not the effect; (C) and (D) don't fit.

6. (B) Insight learning. A sudden solution after a pause, without gradual trial-and-error, is Köhler's insight. (A) and (C) are Tolman's; (D) is gradual operant training — the opposite.

7. (B) Learned helplessness. Repeated uncontrollable shock produced passive failure to escape even when escape was easy. (A), (C), and (D) are conditioning terms that don't describe learned passivity from lack of control.

8. (C). Mirror neurons are real but their role in human imitation and empathy is genuinely debated and frequently overstated — the current cautious view. (A) overstates; (B) is false (they fire during observation too); (D) is false (first reported in monkeys).

9. (C) Vicarious punishment. The employee changes behavior after watching a coworker be punished. (A) requires direct reinforcement of the employee; (B) would require the coworker being rewarded; (D) involves hidden learning shown later, not consequence-driven imitation.

10. (B). The aggressive-model mean (12.4) towers over both the nonaggressive (1.8) and control (2.0) conditions, which are nearly equal — so observing the aggressive model substantially increased imitative aggression. (A) ignores the huge condition difference; (C) is false (1.8 < 2.0); (D) contradicts the data.

11. (B). Improvising a novel route by picturing the city's layout requires a flexible mental representation of space — a cognitive map — not a rote stimulus-response habit. (A), (C), and (D) are all simple conditioning with no spatial representation.

12. (B) Introducing a cash bonus for each time they do the drill. Attaching a salient extrinsic reward to an already-loved activity risks the overjustification effect, eroding intrinsic motivation. (A), (C), and (D) do not bolt a tangible reward onto the behavior and are far less likely to undermine it.

13. (B). Group B's flat-then-plunging curve is the signature of latent learning: the rats had been learning the maze all along, and the introduced reward suddenly motivated performance. (A) misreads the data (they couldn't drop that fast if they were only starting to learn on Day 11); (C) and (D) are the wrong phenomena.

14. (B) Learned helplessness. A capable student who has stopped trying after repeated uncontrollable failure fits learned helplessness exactly. (A) is ruled out because she is described as capable; (C) requires an extrinsic reward undermining intrinsic interest; (D) is a conditioning term, not a description of perceived helplessness.

15. (B). The defining difference: the observational learner acquires behavior by watching a model and the model's consequences, rather than personally performing and being reinforced. (A) is false (observational learning also involves consequences — vicarious ones); (C) is false (animals show observational learning); (D) is false.

AAQ Scoring Rubric (7 points)

| Part | Point(s) | Awarded for |

|---|---|---|

| A | 1 | Naming the method as an experiment (not "random assignment" or "observation"). |

| B | 1 | Operationally defining the free-play DV as the counted number of spontaneous distinctive actions during the alone period. |

| C | 1 | Describing the Reward > Punishment difference in free play and citing the numbers (4.6 vs. 1.3). |

| D | 1 | Identifying a real ethical guideline (informed consent or confidentiality) and describing how it was applied. |

| E | 1 | Committing to a generalizability judgment supported by specific participant evidence (age 4–5, four preschools, one city; balanced sex / mixed ethnicity). |

| F | 2 | (1) Stating that the data support the claim; (2) correctly applying a psychological concept — the learning-vs-performance distinction and/or vicarious punishment / motivation — to the equal prompted scores. |

PsyIQ · Lesson 20 of 30 · Unit 3: Development and Learning. FRQ practice this lesson is an Article Analysis Question (AAQ). Q1-style and AAQ practice 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.

← Lesson 19
Lesson 21 →
Score: 0/0 correct