2.3 Positive and Negative Punishment & Unwanted Effects of Aversives

Functional Definition of Punishment

Within behavior analysis, punishment is defined functionally, not topographically, legally, or morally. Punishment has occurred if and only if a stimulus change immediately follows a response and results in a decrease in the future probability, frequency, rate, or intensity of that response class under similar antecedent conditions (Azrin & Holz, 1966; Skinner, 1953).

Just as with reinforcement, punishment is defined strictly by its empirical effect on future behavior, never by the subjective intent of the change agent, the unpleasantness of the stimulus, or cultural perceptions of penalty. If an angry teacher delivers a loud verbal reprimand, sends a student to the principal's office, or assigns 50 pushups, none of these actions constitute punishment unless objective behavioral data demonstrate a reduction in the future rate of the target behavior. If the student continues or increases the behavior in subsequent sessions, the reprimand functionally operated as reinforcement (e.g., peer attention or escape from academic boredom), regardless of how punitive the teacher intended it to be.

---

Positive Punishment (Type I) vs. Negative Punishment (Type II)

Like reinforcement, punishment is categorized into two basic operant forms depending on whether a stimulus is added or removed contingent on the response:

• Positive Punishment (Type I Punishment): Occurs when a response is immediately followed by the presentation, addition, or increase in intensity of an aversive stimulus, resulting in a decrease in the future frequency of that behavior under similar conditions (e.g., touching a hot stovetop results in an intense thermal burn; future stovetop touching decreases).
• Negative Punishment (Type II Punishment): Occurs when a response is immediately followed by the removal, withdrawal, termination, or reduction of an already present reinforcing stimulus, resulting in a decrease in the future frequency of that behavior under similar conditions (e.g., speeding in a vehicle results in a $250 monetary fine; future speeding decreases).

---

Positive Punishment Procedures in Clinical Practice

Applied behavior analysis utilizes several positive punishment procedures when functionally justified, socially valid, and ethically authorized:

1. Reprimands

Verbal statements of disapproval delivered immediately contingent on problem behavior (e.g., "No, do not touch that"). While commonly used, reprimands are highly susceptible to habituation if delivered frequently. Furthermore, for learners experiencing an establishing operation for social attention, reprimands frequently function inadvertently as positive reinforcement.

2. Response Blocking

Physically intervening as soon as the client initiates problem behavior to prevent the completion of the response. Response blocking is frequently applied to high-risk behaviors such as chronic pica, eye-gouging, or severe self-injurious head-banging.

Critical Distinction: Response blocking must not be confused with operant extinction. In extinction, the client emits the full response, but the reinforcing consequence is withheld (e.g., the child bangs their hand on the table, but social attention is withheld). In response blocking, the physical movement is physically interrupted or prevented from completing.

3. Overcorrection Procedures

Developed by Foxx and Azrin (1972), overcorrection is a behavior-reduction tactic requiring the individual to engage in effortful, relevant physical behavior directly related to the target infraction. Overcorrection encompasses two distinct variations:

• Restitutional Overcorrection: Contingent on problem behavior, the individual must restore the disturbed environment to a condition vastly superior to its state prior to the behavior.
  • Example: A client tips over a desk in an anger outburst. Under restitutional overcorrection, the client must not only right their own desk, but also straighten every desk in the entire classroom, dust the surfaces, and vacuum the classroom rug.
• Positive Practice Overcorrection: Contingent on problem behavior, the individual is required to repeatedly practice the correct, appropriate alternative behavior or topographically correct motor pattern an exaggerated number of times.
  • Example: An elementary student sprints recklessly down the school hallway. Under positive practice, the student is escorted back to the starting point and required to walk slowly and quietly with hands at their sides down the hallway five consecutive times.

4. Contingent Exercise / Effort

Requiring the individual to perform an effortful physical activity (such as doing 20 jumping jacks, running laps, or standing up and sitting down repeatedly) contingent on the occurrence of problem behavior, where the exercise is topographically unrelated to the problem behavior.

---

Negative Punishment Procedures: Response Cost and Time-Out

Negative punishment reduces behavior by removing positive reinforcers or access to reinforcement through two primary methodologies: response cost and time-out from positive reinforcement.

1. Response Cost

Response cost is the contingent loss of a specific quantity of previously earned or established positive reinforcers following the emission of problem behavior (e.g., losing 5 tokens, a deduction of 10 points from a token board, or losing 15 minutes of recess).

Crucial Implementation Guidelines for BCaBAs:

• Clear Behavioral Contingencies: The exact fines and penalized behaviors must be stated explicitly in advance.
• Reinforcer Reserve: The client must have a preexisting bank of earned reinforcers. A client must NEVER go into "token debt" or a negative balance. When a learner accumulates negative points, the value of the token system collapses, leading to profound countercontrol, apathy, and aggression.
• Avoid Excessive Fines: Imposing massive fines suppresses overall responding and damages the reinforcing efficacy of the token economy.

2. Time-Out from Positive Reinforcement

Time-out from positive reinforcement is defined as the contingent withdrawal of the opportunity to access positive reinforcement for a specified duration following the emission of problem behavior.

THE CARDINAL RULE OF TIME-OUT:

Time-in must be demonstrably reinforcing! Time-out is mathematically and functionally defined as the relative contrast between the "time-in" environment and the "time-out" environment. If the instructional environment ("time-in") is non-preferred, boring, chaotic, or contains frustrating academic demands, removing the learner does NOT function as negative punishment. Instead, removal terminates an aversive demand condition, functioning as negative reinforcement (escape)! In such cases, problem behavior will accelerate rather than decline.

Time-out procedures are divided into non-exclusionary and exclusionary formats:

---

Unwanted Side Effects of Punishment and Aversive Procedures

The delivery of aversive stimuli and punishment contingencies generates severe, well-documented behavioral and emotional side effects that every behavior analyst must anticipate and evaluate:

1. Elicited Emotional Responses: Aversive stimuli naturally elicit respondent autonomic reactions, including crying, screaming, panic, sweating, and intense physiological stress.
2. Elicited and Operant Aggression / Countercontrol: As Azrin and Holz (1966) demonstrated, aversive stimulation frequently elicits reflexive aggression toward nearby individuals, objects, or the practitioner delivering the punisher. Operant countercontrol may manifest as physical retaliation, vandalism, or litigation.
3. Escape and Avoidance Behaviors: Through respondent pairing, the change agent delivering punishment, the therapeutic setting, and instructional materials become conditioned aversive stimuli ($S^{av}$). Consequently, clients develop avoidance repertoires, including school truancy, running away (elopement), hiding, and lying.
4. Imitation and Modeling: When authority figures employ coercive, punitive, or aggressive tactics to manage behavior, learners observe and imitate these topographies, utilizing bullying, intimidation, and verbal aggression against weaker peers.
5. Behavioral Contrast: When a response is suppressed under a punishment contingency in one specific stimulus environment (e.g., the therapy clinic), its rate may substantially increase in an unchanged, unpunished environment (e.g., the home or school playground). The client suppresses the behavior only where punishment is actively threatened.
6. The Negative Reinforcement Trap for the Change Agent: Delivering a punisher immediately stops the disruptive or dangerous behavior in real time. Because the cessation of the client's problem behavior is an immediate aversive relief for the teacher, parent, or technician, the change agent's own punishing behavior is powerfully negatively reinforced ($S^{R-}$). This creates a dangerous escalating cycle where practitioners become conditioned to rely exclusively on punishment, increasing its severity when mild punishment habituates, often culminating in systemic abuse.

---

Ethical Mandates and Procedural Protections (BACB Standards)

Due to the significant side effects and human rights concerns surrounding aversives, the Ethics Code for Behavior Analysts (in effect since January 1, 2022) sets the following requirements, supplemented here by widely accepted best practice:

• Prioritize Positive Reinforcement (Standard 2.14): Behavior-change interventions must prioritize positive reinforcement procedures (e.g., DRA, DRI, DRO, noncontingent reinforcement, functional communication training).
• Restrictive or Punishment-Based Procedures Last (Standard 2.15): Behavior analysts recommend and implement restrictive or punishment-based procedures only after less intrusive means have not produced the desired results, or when an existing intervention team determines that the risk of harm from the behavior outweighs the risk of the procedure. Best practice is to pair any punishment procedure with reinforcement for an acceptable alternative behavior, so the learner has another way to access reinforcement.
• Assessment First (Standards 2.13 and 2.14): Interventions are based on assessment results, so the function of the behavior should be identified (for example, through a functional behavior assessment) before consequences are chosen.
• Consent, Review, and Monitoring (Standards 2.11 and 2.15): Obtain informed consent, comply with any required review process (for example, a human rights review committee), and continually evaluate and document effectiveness, modifying or discontinuing an ineffective restrictive procedure in a timely manner.

---

Comparative Matrix of the Four Basic Operant Contingencies

---

Common BCaBA Exam Traps: Punishment and Aversive Control

• Trap 1: Assuming a Nominal Penalty is Punishment: Just because an intervention is labeled a "detention," "scolding," or "time-out" does not make it punishment. If the target behavior does not decrease in future frequency, it is NOT punishment.
• Trap 2: Time-Out Without Functional Time-In: Implementing time-out for a behavior maintained by escape (such as refusing to complete a difficult task) is a classic exam blunder. Sending the learner away provides negative reinforcement, accelerating the behavior.
• Trap 3: Token Debt in Response Cost: Permitting clients to accumulate negative balances in token economies damages stimulus control, induces learned helplessness, and triggers countercontrol. Response cost must only withdraw previously earned tokens up to a zero balance.