Reinforcement, Schedules and Learned Reinforcers

Reinforcement is what happens when a consequence makes a behavior more likely to happen again. As a technician, you deliver reinforcement exactly as your supervisor's written plan describes. You also collect the data that shows whether it is working.

By the end, you should be able to:

• Classify a procedure as positive or negative reinforcement from what was added or removed and what the data show.
• Deliver a reinforcer immediately and only after the target behavior.
• Follow continuous, ratio and interval schedules, including when an interval reinforcer is actually earned.
• Pair a new reinforcer with an established one as your plan directs.
• Recognize an ineffective reinforcer or satiation and report it.

Before you start: No prior knowledge is needed. Each term is explained when it first appears.

What Makes Something Reinforcement

Three words carry this lesson:

• A behavior is something a person does that you can observe and count.
• A consequence is what happens right after the behavior.
• Future likelihood is how often the behavior happens on later opportunities.

Reinforcement means a consequence follows a behavior, and that behavior then happens more often in the future. In plain terms, something counts as reinforcement only if the behavior goes up afterward. Your intentions do not decide it, and neither does how pleasant the consequence seems.

Example: Rafi's plan targets saying "my turn" during a board game. Each time Rafi says it, the technician hands over the dice. Data across many turns show "my turn" happening more often. That repeated increase is why the dice access is called a reinforcer.

One instance proves little, because a single "my turn" could happen for many reasons. A pattern across repeated opportunities is stronger evidence, and one session can contain many opportunities. Your supervisor interprets that pattern. Your job is accurate data.

Tempting mistake: "A reward must reinforce." Suppose a plan gives a sticker after each completed worksheet row. Across many opportunities, completed rows stay flat. The sticker was meant as a reward, but the data show it is not reinforcing that behavior. Record what happened and tell your supervisor.

Positive or Negative: Added or Removed

Both kinds of reinforcement increase behavior. "Positive" and "negative" do not mean good and bad. They mean added and removed.

• Positive reinforcement: something is added after the behavior, such as dice, praise or a token. The behavior then increases.
• Negative reinforcement: something is removed after the behavior, such as a task. The behavior then increases.

Example: Later in the session, Rafi works on a sorting task. The plan says that when Rafi says "break, please," the technician immediately removes the sorting materials for the planned break. Across many opportunities, Rafi's requests increase. Something was removed, and the behavior increased. That is negative reinforcement.

Change one fact: The materials are removed the same way, but requests become less frequent over time. This is no longer reinforcement, because the behavior did not increase.

Tempting mistake: "Negative reinforcement is punishment." Punishment decreases behavior. Negative reinforcement increases it. To classify any procedure, ask two questions:

• What was added or removed?
• Did the behavior go up or down over repeated opportunities?

You never create an unpleasant situation just so you can remove it. You remove only what the plan specifies, when the plan specifies.

Checkpoint 1

(a) In a group activity, Dana's plan says that when Dana says "too loud," the technician turns the music down. Across many group sessions, Dana says "too loud" more often. Which type of reinforcement is this?

(b) Another learner gets a high five after each cleanup. Across two weeks of opportunities, cleanups have not increased. Is the high five a reinforcer for cleaning up?

Answer: Checkpoint 1

(a) This is negative reinforcement. The loud volume was removed, and the request increased. It is not punishment, because the behavior went up, not down.

(b) Not according to these data. A high five is common praise, but praise does not reinforce everyone's behavior. Keep following the plan, record accurately, and report the flat data to your supervisor. Do not swap in a different reinforcer yourself.

Delivering Reinforcement as Planned

• Contingent means you deliver the reinforcer only after the target behavior, not at other times.
• Immediate means you deliver it right after the behavior, before something else happens in between.

A delay can cause problems. Suppose Rafi says "my turn," then grabs another player's piece, and only then gets the dice. The dice now follow the grabbing.

Plans also set the reinforcer's dimensions, which describe how the reinforcer is delivered. The descriptions below are short everyday explanations to help you picture each one. They are not quotations or official BACB definitions, and your plan states the exact details:

• Magnitude is how much or how long, such as 30 seconds versus 2 minutes of a game.
• Intensity is how strong the reinforcer is, such as a light-up toy on its dim setting versus its bright setting.
• Variety is rotating among different reinforcers rather than giving the same one every time.

More or stronger is not automatically better. You deliver the dimensions the plan states. You do not adjust them on your own, and you never make anything unpleasant stronger.

Schedules: Which Responses Earn Reinforcement

A schedule of reinforcement is the rule for which responses earn the reinforcer. Each response is one instance of the target behavior.

• Continuous reinforcement: every target response earns it. Plans commonly use this while a learner is first learning a skill.
• Intermittent reinforcement: only some responses earn it. After the learner meets a plan criterion, a supervisor may "thin" reinforcement by moving to an intermittent schedule.

Intermittent schedules come in two families:

• Ratio schedules count responses.
  • Fixed ratio: after a set number of responses, such as every 3rd correct response.
  • Variable ratio: after a number that varies around an average. For an average of 3, sometimes the 2nd response earns it and sometimes the 4th.
• Interval schedules use time, but they still require a response.
  • Fixed interval: once a set amount of time has passed, the first target response after that point earns the reinforcer.
  • Variable interval: the same rule, but the amount of time varies around an average.

Example: Rafi's plan uses a 2-minute fixed interval for correct sorting. When 2 minutes pass, nothing is delivered yet. The next correct placement earns the token, and you deliver it immediately. Your plan states when the timing starts.

Tempting mistake: "The timer ended, so deliver." Time passing alone does not earn an interval reinforcer. If Rafi is not sorting when the time is up, you wait for the next correct placement. Delivering something because the clock ran out is a different, time-based procedure. It is not interval reinforcement.

No schedule is always best. Your supervisor chooses the schedule, and you implement it exactly as written. If you think a different schedule might work better, tell your supervisor rather than switching.

Checkpoint 2

(a) Dana's plan uses a 3-minute fixed interval for writing correct answers. The 3 minutes end while Dana is looking out the window. A coworker says, "Time's up, give the token." What do you do?

(b) Dana's plan says every 4th correct answer earns a token. You suspect varying the count would work better. What do you do?

Answer: Checkpoint 2

(a) Wait. The token is earned by the first correct answer after the interval ends, so deliver it immediately when Dana writes the next correct answer. Giving the token just because time passed would not follow the fixed-interval plan; it would make delivery time-based rather than contingent on the next correct answer. You cannot say it reinforced looking out the window unless later data show that looking increased.

(b) Keep using every 4th correct answer, which is a fixed ratio. Share your idea with your supervisor. Technicians do not change schedules on their own.

Learned Reinforcers

A conditioned reinforcer, also called a secondary reinforcer, is something that became effective because it was paired with an already effective reinforcer. Praise and tokens are common examples. They are not automatically effective for everyone. For a learner who has never had praise paired with anything, "great job" may change nothing.

Pairing: A plan may direct you to say "great job" just before delivering an established reinforcer, so the praise comes to signal it.

Exchange: Tokens get their value from being traded for backup reinforcers, which are items or activities the learner can buy with earned tokens.

Example: Rafi earns a token for each correct placement and trades 5 tokens for 2 minutes of the dice game. Over sessions, correct placements rise. The tokens are working as reinforcers.

The data show whether pairing has worked: the behavior should increase when you deliver the conditioned reinforcer. If pairing or exchange stops, a conditioned reinforcer can lose its value. That is why plans keep the pairing or exchange going.

Checkpoint 3

Dana's data already show that Dana works to earn 20 seconds of a favorite song. Dana's plan now directs pairing: after each correct answer, say "nice writing," then immediately play the song for 20 seconds.

(a) In what order do you deliver the praise and the song? A coworker starts saying "nice writing" at random times during the session, with no song afterward, because "praise is always good." Is the coworker pairing?

(b) Several weeks later, the plan has you give "nice writing" alone on some answers to check whether the praise now works by itself. On those answers, correct answers have not increased. A teammate suggests you start pairing the praise with a snack instead. What do you do?

Answer: Checkpoint 3

(a) Say "nice writing" first, then play the song right away, and only after a correct answer. The praise has to come just before the established reinforcer so it can come to signal it. The coworker is not pairing. The praise is not followed by the song, and it is not given after the target behavior. Repeated praise without the song can also weaken the very link the plan is trying to build. Report the inconsistent delivery to your supervisor.

(b) Keep following the plan as written, record the data accurately, and report to your supervisor that praise alone has not increased correct answers. Switching to a snack would be redesigning the procedure, which is your supervisor's decision. The flat data tell you the praise is not yet working as a reinforcer for this behavior; they do not tell you what to change.

Preference, Satiation and Reporting

Preference is what someone chooses or spends time with. A preferred item is a good candidate reinforcer, but it is proven only when the behavior that earns it increases. A learner may happily choose an item and still not work to earn it.

Satiation happens when a reinforcer loses its motivating value, often because it was used too frequently. Signs include the learner choosing the item less often or showing less engagement or interest in it. Plans may prevent satiation with reinforcer menus or rotation.

Your role:

• Deliver reinforcement exactly as written.
• Record data accurately.
• Promptly report to your supervisor when:
  • the behavior is not increasing,
  • you see signs of satiation,
  • team members are delivering reinforcement differently, or
  • something might affect progress, such as illness or a schedule change.

Recap

• Reinforcement is defined by a future increase in behavior, shown across repeated opportunities. It is not defined by your intent, by how pleasant something seems, or by what happened to come right before it.
• Positive reinforcement adds something; negative reinforcement removes something. Both increase behavior. Punishment decreases it.
• Deliver reinforcement immediately and only after the target behavior, with the magnitude, intensity and variety the plan specifies. More or stronger is not automatically better.
• Ratio schedules count responses. Interval schedules require a response after the time passes.
• Conditioned reinforcers gain value through pairing or exchange and must be maintained. Pair in the order the plan directs: the new reinforcer just before the established one, after the target behavior.
• Report ineffective reinforcers and satiation. Never change the plan yourself.

Sources

• Sam, A., & AFIRM Team. (2024). Reinforcement, Updated. UNC Frank Porter Graham Child Development Institute. https://afirm.fpg.unc.edu/wp-content/uploads/Reinforcement-Brief-Packet-Sam-AFIRM-Team-Updated-2024.pdf (pp. 4, 14, 26, 29–32, 42)
• Behavior Analyst Certification Board. (2021). RBT Ethics Code (2.0). https://www.bacb.com/wp-content/rbt-ethics-code (2.02, 2.05, 2.06)
• Behavior Analyst Certification Board. (2023). RBT Test Content Outline (3rd ed.). https://www.bacb.com/wp-content/rbt-outline-3rdEd/
• Vanderbilt IRIS Center. Behavior and Classroom Management, Page 5: Reinforcement (features table, magnitude). https://iris.peabody.vanderbilt.edu/module/bp/cresource/q1/p05/
• Trosclair-Lasserre, N. M., et al. (2008). Reinforcement magnitude: An evaluation of preference and reinforcer efficacy. https://pmc.ncbi.nlm.nih.gov/articles/PMC2408337/ (introduction, as indexed; cited for the distinction between quantity, intensity and duration)
• Effects of neutral and enthusiastic praise on the rate of discrimination acquisition. https://pubmed.ncbi.nlm.nih.gov/29446093/ (indexed record)
• The effects of constant versus varied reinforcers on preference and resistance to change. https://pmc.ncbi.nlm.nih.gov/articles/PMC2861876/ (indexed abstract)
• Shull, R. L. (1970). A response-initiated fixed-interval schedule of reinforcement. J Exp Anal Behav, 13(1), 13–15. https://pmc.ncbi.nlm.nih.gov/articles/PMC1333651/ (laboratory study with pigeons; cited for interval-schedule terminology only)
• NIST OSAC Lexicon: "Conditioned Reinforce" and "Variable Interval Schedule of Reinforcement." https://www.nist.gov/glossary-term/20881 ; https://www.nist.gov/glossary-term/34346 (canine detection glossary; cited for terminology only)
• Vanderbilt TRIAD. Preference Assessments: An Introduction for Educators. https://vkc.vumc.org/assets/files/triad/tips/preference_assessment.pdf (p. 7)

Independent study resource

This study resource is not affiliated with, sponsored by, or endorsed by the Behavior Analyst Certification Board (BACB). BACB, Registered Behavior Technician and RBT are registered trademarks of the Behavior Analyst Certification Board. Source citations do not imply endorsement by their authors or publishers.