Instructional Control, Reinforcement & Punishment, and Extinction

Instructional Control

  • Instructional control is the likelihood that instructions elicit a correct response from the learner, or building a history of reinforcing compliance.
  • Occurs quicker for some clients versus others.
  • To gain instructional control, pairing must occur first.
  • Consistency is key: Follow through with instructions and expectations. Our words have meaning.
  • Demands should be faded in, following a schedule set by the BCBA.
    • Start with previously mastered targets.
    • Provide reinforcement.
    • Fade in targets that are in the acquisition phase.
    • This process varies for each learner.
  • To fade in demands:
    • Begin by playing and pairing with the learner.
    • Give one simple, known, or mastered demand.
    • Use behavioral momentum.
    • If the learner complies, provide reinforcement.
    • If the learner does not comply, prompt the correct response as written in the patient's ITP, then repeat the demand to get an independent response.
    • Increase to one mastered demand followed by one targeted demand or new skill.
    • If the learner complies, provide reinforcement.
    • If the learner does not comply, prompt the correct response as written in the patient's ITP.
    • Increase the number of demands before providing reinforcement.
    • See patient's ITP and BCBA for specific details and optimal schedule of reinforcement.
  • Having instructional control does not automatically mean you are paired.
    • Aversive control (drill sergeant effect): Learner responds to avoid trouble, not for reinforcement.
  • Pairing throughout the entire therapeutic relationship is crucial.
  • Being paired does not guarantee instructional control.
    • Big brother effect: Fun to be around, but compliance is unlikely.
  • Ideally, there should be a balance between pairing and instructional control.
  • To teach effectively, both are needed.
  • Seven steps to earning instructional control (Robert Schrum):
    1. Maintain control of reinforcers: Restrict access to preferred items.
    2. Make interactions enjoyable: Pair to make the child want to listen.
    3. Say what you mean and mean what you say: Be trustworthy; follow through with directions.
    4. Reinforce appropriate behavior: Show the child it's beneficial to follow directions; reinforce compliance frequently.
    5. Initially reinforce after every response, then thin out reinforcement variably (guided by BCBA).
    6. Prioritize goals: Track the child's reinforcers and align with what's important to them and your goals.
    7. Withhold reinforcement for inappropriate behavior: Reinforce positive behaviors, not maladaptive ones.
  • When shadowing, observe:
    • How instructional control was gained.
    • The instruction given.
    • Whether the learner followed the instruction the first time.
    • What happened after the learner followed the instruction.

Reinforcement and Punishment

  • Reinforcement: A stimulus that is added or removed to increase the future occurrence of a target behavior.
  • Punishment: A stimulus that is added or removed to decrease the future occurrence of a target behavior.
  • Objectives: Understand the difference between reinforcement and punishment, and understand the four methods to reinforcement effectiveness (DISC).
  • Whether positive or negative, reinforcement will always increase a behavior.
  • Positive Reinforcement: The contingent presentation of a stimulus immediately following a response that increases the likelihood of the response occurring again in the future.
    • Example: Mother gives praise for doing homework, increasing future homework behavior.
    • Example: Baby cries, father gives milk; milk is added (positive), increasing future crying (reinforcement).
  • Negative Reinforcement: The contingent removal, termination, or postponement of a stimulus that increases the response in the future.
    • Example: Jason washes dishes to stop mom's nagging; nagging is removed (negative), increasing dishwashing (reinforcement).
    • Example: Amy eats broccoli to leave the dinner table; dinner removal (negative) increases broccoli eating (reinforcement).
  • Unconditioned Reinforcers (Primary/Unlearned): Stimuli that are naturally reinforcing; basic needs such as food, water, air, shelter, safety.
  • Conditioned Reinforcers (Secondary/Learned): Stimuli that are learned through interactions with the environment; vary and change over time (e.g., money, tokens, good grades, praise).
  • Principles of Reinforcement Effectiveness (DISC):
    • Deprivation: Decrease in access to a reinforcer increases its effectiveness.
      • Example: Limited Snickers access makes them more reinforcing.
    • Satiation: Increase in access to a reinforcer decreases its effectiveness.
      • Example: Offering PlayStation after Chris played video games all night will not be an effective reinforcer.
    • Immediate: Reinforcer must be delivered immediately following the target response.
      • Faster delivery is more effective.
      • Delays can lead to inadvertent reinforcement of a different behavior.
    • Size: The amount or size of reinforcement must be large enough and appropriate for the activity.
      • Too small = low motivation; too large = satiation.
      • Example: One math problem = half a Skittle; 30 math problems = handful of Skittles.
      • Work to play ratio needs to be balanced.
    • Contingency: Reinforcers are presented when the correct target behavior occurs; withheld for undesired behavior.
      • Use "first, then" language to establish contingencies.
      • Example: "First, finish one math problem, then you get a Skittle."
  • Considerations during reinforcement:
    • Many things can be reinforcing (tangible items, praise, physical interaction, edibles, playtime, etc.).
    • What is reinforcing to one child may be aversive to another.
    • Continually assess preferences and individualize reinforcers; preferences change often.
    • Obtain approval from parent/caregiver and BCBA before using edible reinforcers.
    • Assess for allergies or dietary preferences.
    • Food from other patients or the RBT should never be shared without caregiver consent.
    • Socially reinforce the specific behavior (e.g., "I love how you follow directions" instead of just "good job").
    • Use behavior-specific praise: "You're right, those are both cows".
    • Ensure vocal praise is appropriate for the learner's age.
    • Tell the child what you do want them to do rather than what you don't want them to do (e.g., "walk, please" instead of "don't run").
    • Use accessing reinforcement as a learning opportunity; expand on the target with social praise (e.g., "That is a cow. A cow says moo").
    • Be mindful of tone of voice (instructional vs. praise).
    • Often the best reinforcer is yourself (successful pairing).
  • Punishment: A stimulus that is either added or removed to the environment and decreases the target behavior from occurring in the future.
  • Whether positive or negative, punishment will always decrease a behavior.
  • Positive Punishment: The contingent presentation of a stimulus immediately following a response that decreases the occurrence of the behavior in the future.
    • Example: Teacher reprimands Carly for picking her nose, and she stops. The reprimand is added (positive), decreasing nose-picking (punishment).
  • Negative Punishment: The contingent removal of a stimulus immediately following a response that decreases the behavior occurring again in the future.
    • Example: Lauren kicks a peer and is removed from recess. Recess is removed (negative), decreasing kicking (punishment).
  • Considerations of Punishment:
    • Often used in the form of negative punishment (e.g., response cost - losing a token).
    • If a punishment procedure is in place, a reinforcement procedure must also be used (e.g., gain tokens for positive behavior).
    • Negative side effects: anger, retaliation, damaged relationship between staff and learner.
    • Behavioral contrast: Learner only understands the punishment with the person who taught them, not with other staff or parents.
    • If not used consistently, the behavior can strengthen or increase.
    • BCBAs are ethically responsible to use punishment carefully, sparingly, and only when other interventions have failed.
    • Decision to use punishment may only come from BCBA after consent from parent/caregiver and senior clinician.
    • RBTs should never use punishment without consent and training from the BCBA.
    • Punishment procedures are very rare in ABA; focus is commonly on reinforcement.

During shadowing, observe:
* If reinforcement is being used; practice using reinforcement.
* If learner received reinforcement or punishment, and whether it was positive or negative.

Extinction

  • Extinction: When reinforcement for a previously reinforced behavior is withheld, terminated, or removed.
  • A consequence that maintains a behavior is removed.
  • Unethical to use extinction for aggression or self-injurious behaviors, as it can temporarily increase these behaviors.
  • Extinction Burst: The temporary increase in the rate of a behavior following implementation of an extinction procedure.
    • Example: Child yells a curse word and parents laugh. They stop laughing, and the child initially increases the frequency and intensity of saying the word to gain attention, then stops altogether.
  • When planning to extinguish a behavior, teach a replacement behavior.
    • Example: Teach the child to ask for attention or tap someone on the shoulder instead of yelling a curse word.
    • The replacement behavior needs to be easier to engage in than the problem behavior it is replacing.
  • Spontaneous Recovery: The reappearance of the extinguished behavior after a period in which the behavior has not been reinforced.
  • Graph:
    • Original behavior response frequency (initial behavior)
    • Vertical line = reinforcement removed (extinction procedure implemented).
    • Spike in behavior = extinction burst.
    • Behavior decreases as the extinction procedure works.
    • Small jump in behavior toward the end of the graph = spontaneous recovery.
    • If the extinction procedure continues, the behavior will quickly decrease again.