chapter 10

avoidance (active avoidance)

Avoidance (negative reinforcement) involves performing a behavior to prevent an aversive stimulus. It increases operant responding.
Example: 

A rat presses a lever to avoid a shock.

A person takes medicine to avoid getting sick.

punishment (passive avoidance)

Punishment involves refraining from a behavior to avoid an aversive consequence. It decreases operant responding.
Example: 

A child avoids touching a hot stove after being burned once.

A student stops talking out of turn after receiving detention.

Why is Bechterev’s 1913 experiment not a true example of Pavlovian conditioning?

Although it involved a warning stimulus (CS) followed by a shock (US), participants learned to actively respond (lifting their finger) to avoid the shock.
This makes it instrumental/operant conditioning, not Pavlovian, because the response was learned to control the outcome (avoid the US).

Example:
Participants learned to lift their fingers during the CS to prevent shock—not just passively associate CS and US.

How did Brogden et al. (1938) demonstrate that avoidance conditioning differs from classical conditioning?

They tested guinea pigs in a rotating wheel using a tone (CS) and shock (US):

• Classical group: Shock followed the tone no matter what.

• Avoidance group: If animals moved the wheel during the tone, the shock was omitted.

Key Finding:

• The avoidance group learned two things:

  1. Pavlovian learning – the tone predicts a shock.

  2. Instrumental learning – moving prevents the shock.

• Conclusion: Avoidance conditioning involves active behavior to prevent the US, making it fundamentally different from standard classical conditioning.

According to the Two-Process Theory, how is avoidance behavior learned and reinforced?

Avoidance behavior isn’t reinforced by escaping the aversive event—it’s reinforced by terminating the warning signal(CS).

Process 1: Classical Conditioning

• The warning signal (e.g., tone) becomes aversive by being paired with the shock (fear conditioning).

Process 2: Operant Conditioning

• The animal learns that making a specific response ends the warning signal, reducing fear.

• This negative reinforcement maintains the avoidance behavior.

Example: A rat hears a tone (CS) that predicts shock. It learns to jump to another compartment when the tone starts—not to avoid the shock directly, but to stop the tone and reduce fear.

How do acquired-drive experiments support the Two-Process Theory of avoidance?

They separate classical and operant learning to test if both are necessary and sufficient for avoidance learning.

Problem:
In typical avoidance procedures, fear learning (classical) and response learning (operant) happen together—making it hard to isolate them.

Experimental Design:

1. Phase 1 – Classical Conditioning:

   • The warning signal (CS) is paired with the shock (US) to induce fear.

2. Phase 2 – Escape Training:

   • The CS (now feared) is presented, and the subject learns to make a response to terminate the CS, reducing fear.

Result:
Subjects learn to escape the CS, supporting that:

• Fear of the CS is learned via classical conditioning.

• The escape response is reinforced through fear reduction—matching the Two-Process Theory.

Why is it a problem for the Two-Process Theory that fear and avoidance are not always positively correlated?

The theory predicts:

More fear → stronger avoidance

But in reality:

• As animals get better at avoiding shocks, their fear decreases

• Kamin, Brimer, & Black (1963):

  • Animals trained to avoid shocks 27 times in a row showed less conditioned fear than those with only 9 avoidances

  • Fear was measured via conditioned suppression (less fear = less suppression)

✅ Conclusion:
Avoidance behavior continues even when fear goes down, contradicting the theory's assumption that fear reduction is the main reinforcer.

How does unsignaled avoidance (sidman) contradict the Two-Process Theory?

The Two-Process Theory requires a warning signal (CS) to induce fear.
But in Sidman (unsignaled) avoidance, there is no CS:

• Shocks are delivered every few seconds (S–S interval), unless an animal makes a response to delay it (R–S interval)

• Despite the lack of warning, animals still learn to avoid shocks

✅ Conclusion:
Avoidance occurs without any conditioned fear stimulus, showing that classical fear conditioning isn’t always required.

What’s the extinction problem in the Two-Process Theory of avoidance?

The theory predicts a yo-yo pattern:

• As fear goes down (no shocks), avoidance should extinguish

• When shock returns, fear increases and avoidance resumes

📉 But in practice:

• Avoidance persists even when no shocks are presented for long periods

• Fear decreases, yet animals continue responding

✅ Conclusion:
Avoidance behavior is resistant to extinction, suggesting it may be maintained by factors other than fear reduction, like habit formation or cognitive expectations.

How can avoidance behavior be extinguished according to expectancy theory?

Avoidance persists because subjects form two expectancies:

1. Stimulus-Outcome: The warning signal (CS) predicts shock if no response is made.

2. Response-Outcome: Making the avoidance response leads to safety (no shock).

🔁 These beliefs maintain avoidance because:

• Expectancy #1 is never tested (they never get shocked because they always respond)

• Expectancy #2 keeps getting reinforced (response always leads to no shock)

🛑 To extinguish avoidance:
You must break these expectancies.
One method:
Response Prevention (Flooding)

• Physically prevent the avoidance response during the warning signal

• The subject learns: the shock doesn’t happen even if they don’t respond

• New learning occurs: "I’m safe without responding"

✅ Result: Expectancies are weakened, and avoidance behavior decreases.

What punishment intensity is most effective?

Start with a high intensity punishment right away.
Example: A firm "No!" the first time a dog bites is more effective than gradually increasing your reaction.

When should punishment be delivered to be most effective?

Immediately after the behavior, and it must be contingent on the behavior.
Example: Spraying a cat with water while it’s scratching, not after.

Which schedule makes punishment more effective?

Use a continuous schedule—punish every instance of the behavior.
Example: Stop a child from yelling every time they do it, not just sometimes.

How does offering alternatives affect punishment?

Punishment is more effective when the subject has a clear alternative to get the reward.
Example: Give a child a toy to fidget with instead of punishing them for tapping the table.

What’s the effect of slowly escalating punishment?

Slow escalation builds resistance. It’s better to start strong.
Example: Don’t go from time-outs to grounding—start with grounding if needed.

What happens if you start with severe punishment and then reduce it?

The effect of the severe punishment persists, making even mild punishment effective later.
Example: One harsh consequence early on can make a teen comply even with smaller consequences later.

Should punishments be reused across behaviors?

A: No—punishments are behavior-specific.
Example: If a consequence was used for yelling, it won’t automatically work for lying.

ways to make punishment effective (6)

• As intensity of punishing stimulus increases, degree of suppression increases.

• Works best when contingent on the response and not delayed. [bitter apple spray; stay home and catch them in the act.

• Works best if continuous rather than partial reinforcement schedule. (i.e. the degree of response suppression produced by punishment depends on the proportion of responses punished)

• If you give organism an alternative, an unpunished route to reinforcement, then effects of punishment are enhanced

• Don’t do a bunch of small punishments then a large one. Would be better to do large and then some smalls. (Yell loudly first time)

• Use a punishment specific for that offense. Punishments don’t generalize. It might be hard for the animal to learn that some already used punishment is now associated with some new behavior you are trying to extinguish.