Study guide · Area G

Behavior-Change Procedures

25 of the exam's 175 scored questions (14%) · 19 tasks

This area covers the procedures behavior analysts use to change behavior: reinforcement, differential reinforcement, time-based schedules, conditioned reinforcers, antecedent variables, discrimination training, prompting and fading, modeling, instructions, shaping, chaining, trial-based and free-operant teaching, group contingencies, generalization and maintenance, punishment, emotional side effects and emergent relations. Questions give a case and ask which procedure to use or change, or what went wrong; the right answer follows the principle behind the procedure and the function of the behavior.

A study feature written by ExamDragon, not part of the real exam. Each heading is one of the BACB's tasks for this content area (19 of them), in the BACB's words.

G.1Design and evaluate positive and negative reinforcement procedures.

Positive reinforcement procedures deliver a stimulus contingent on the target response; negative reinforcement procedures remove or postpone a stimulus (a demand, noise, an aversive task) contingent on it.
Design steps: define the target behavior, identify reinforcers through preference and reinforcer assessments (or the behavior's function), set the contingency and schedule, deliver the consequence immediately and consistently, and pair it with praise.
Effectiveness depends on immediacy, contingency (the reinforcer is given only after the target behavior), magnitude, quality, the relevant motivating operation and the schedule (CRF for acquisition, intermittent for maintenance).
Negative reinforcement can be used constructively: a break from work contingent on completing a set number of problems, or the removal of a nonpreferred task when the learner asks appropriately.
Evaluate with data: if the target does not increase, check the reinforcer's value, the delay, the schedule, competing sources of reinforcement, response effort and procedural integrity.
Plan to thin the schedule and shift to natural reinforcers once the behavior is established.

In practice: A student rarely completes math work. After a preference assessment, the behavior analyst arranges for every completed row of problems to earn a break of 2 minutes with drawing materials (negative reinforcement by removal of work plus positive reinforcement by access to drawing). Completion rises; the requirement is gradually increased.

Where questions turn: If the behavior does not increase, the consequence was not a reinforcer under those conditions; the procedure, not the learner, needs to change.

Practise G.1

G.2Design and evaluate differential reinforcement (e.g., DRA, DRO, DRL, DRH) procedures with and without extinction.

Differential reinforcement of alternative behavior (DRA) reinforces a specific alternative response and withholds reinforcement for the problem behavior; differential reinforcement of incompatible behavior (DRI) is a form of DRA in which the alternative cannot occur at the same time as the problem behavior.
Differential reinforcement of other behavior (DRO) delivers reinforcement when the problem behavior has not occurred for a specified interval (fixed or variable, resetting or not) or at a specified moment (momentary DRO); intervals start short, based on baseline IRT, and are lengthened.
Differential reinforcement of low rates (DRL) reinforces responding at a lower rate, used when the behavior is acceptable at low levels: full-session DRL (reinforcement if total responses are at or below a limit), interval DRL, and spaced-responding DRL (reinforcement when an IRT exceeds a minimum).
Differential reinforcement of high rates (DRH) reinforces responding at or above a criterion rate, used to increase behaviors that occur too slowly.
With extinction, the problem behavior no longer produces its reinforcer; this usually makes DR procedures more effective. Without extinction, DR can still work if the alternative produces reinforcement that is denser, of higher quality, more immediate or requires less effort than the problem behavior.
Use the function: the reinforcer for the alternative behavior should match the reinforcer that maintained the problem behavior, and the alternative should be easy and efficient.

In practice: A boy hits to get his teacher's attention. The teacher gives enthusiastic attention every time he taps her arm and says excuse me, and hitting now produces no attention (DRA with extinction). In a different case, a student who talks at a reasonable rate but too often earns a reward if he contributes 5 times or fewer in a class (full-session DRL).

Where questions turn: DRO reinforces the absence of the problem behavior for an interval, whatever else the person is doing; DRA reinforces a specific alternative behavior.

Practise G.2

G.3Design and evaluate time-based reinforcement (e.g., fixed-time) schedules.

Time-based schedules deliver a stimulus on a fixed-time (FT) or variable-time (VT) schedule independent of behavior; noncontingent reinforcement (NCR) is the use of such schedules with the reinforcer that maintains problem behavior.
NCR reduces problem behavior mainly by abolishing the motivating operation (the person has free access to the reinforcer) and by breaking the contingency between behavior and reinforcer (extinction, when used).
Design steps: identify the maintaining reinforcer from a functional assessment, start with a dense schedule (often based on baseline IRT, so the reinforcer comes more often than the behavior usually does), then thin the schedule gradually as behavior stays low.
Advantages: easy to implement, creates an enriched environment, reduces the likelihood of an extinction burst; risks: adventitious reinforcement (the reinforcer happens to follow problem behavior and strengthens it) and reduced motivation for adaptive responses that produce the same reinforcer.
Evaluate by the reduction in problem behavior and by whether it persists as the schedule is thinned; NCR is often combined with teaching an alternative response.

In practice: A woman screams to get staff attention. Staff deliver brief attention every 30 seconds regardless of her behavior (FT 30 s), and screaming drops; over several weeks they lengthen the interval to 5 minutes while teaching her to ask for a conversation.

Where questions turn: FT and VT schedules require no response for delivery, unlike FI and VI schedules, which require a response after the interval ends.

Practise G.3

G.4Identify procedures to establish and use conditioned reinforcers (e.g., token economies).

Conditioned reinforcers are established by pairing a neutral stimulus with existing reinforcers, presenting it just before or with the reinforcer, until it acquires reinforcing value.
Test the new conditioned reinforcer by making it contingent on a response and checking whether the response increases; continued pairing keeps it effective.
A token economy has defined target behaviors, tokens (conditioned generalized reinforcers), backup reinforcers, exchange rates, and rules for when and where tokens are earned and exchanged.
Start with immediate delivery and frequent exchanges, with small token requirements, then increase the requirement and delay exchange as performance becomes stable; teach the value of tokens through pairing if needed.
Response cost (losing tokens for problem behavior) can be added, but it must not leave the person with a balance so low that the system stops working.
Plan to fade the token system toward natural reinforcers (praise, grades, privileges), for example by lengthening intervals, reducing tokens, and delivering praise with every token.

In practice: A classroom token economy gives students a point each time they follow a direction, exchanged at the end of the day for items from a menu. For a new student who does not yet value points, the teacher first gives a point and immediately lets him exchange it for a snack several times, establishing the points as reinforcers.

Where questions turn: Tokens are conditioned reinforcers only if they are paired with or exchangeable for backup reinforcers; tokens that never buy anything stop working.

Practise G.4

G.5Incorporate motivating operations and discriminative stimuli into behavior-change procedures.

Antecedent interventions alter motivating operations or discriminative stimuli before behavior occurs, often reducing problem behavior and increasing desired behavior without changing consequences.
Using MOs: create establishing operations to increase motivation for target behaviors (limit access to a reinforcer before teaching mands for it), and abolishing operations to reduce motivation for problem behavior (noncontingent attention, reducing task difficulty, offering choices).
Using SDs: present clear discriminative stimuli for desired behavior (a visual schedule, a cue card), and remove or alter SDs for problem behavior (removing items that signal reinforcement for problem behavior).
Examples of antecedent strategies: high-probability request sequences, choice making, task interspersal, environmental enrichment, curriculum revision, priming, and functional communication with signals for when reinforcement is available.
Antecedent changes rarely work alone over time; combine them with consequence strategies (reinforcement of alternatives, extinction) and fade them when possible.

In practice: A child grabs toys from peers. Before free play, the teacher gives him ample individual attention (reducing motivation to seek it through grabbing) and puts a visual card showing ask on the table as a cue for requesting, then reinforces requests.

Where questions turn: Changing an antecedent's value (an MO) differs from changing what it signals (an SD), and both differ from changing consequences, which only affect future behavior.

Practise G.5

G.6Design and evaluate procedures to produce simple and conditional discriminations.

Simple discrimination: one stimulus controls the response; the response is reinforced in the presence of the SD and not in the presence of S-deltas (say red when red card is shown).
Conditional discrimination: the function of a stimulus depends on another stimulus; in matching to sample the sample determines which comparison is correct (touch the picture matching the spoken word).
Design: use differential reinforcement, clear SDs, and arrangements that prevent faulty stimulus control: counterbalance comparison positions, rotate targets, avoid cues from the instructor, and require an observing response to the sample when needed.
Current recommendations favor teaching conditional discriminations as conditional from the start (sample presented first, several comparisons, rotation) rather than teaching a simple discrimination first and expanding.
Evaluate with data on correct responding for each stimulus and position, and probe for errors that suggest faulty stimulus control (position bias, choosing the last reinforced item).

In practice: To teach a child to point to animals by name, the behavior analyst presents three pictures, says touch cow, and rotates positions and targets across trials. Errors cluster on the left position, so she checks for position bias and adjusts the arrangement.

Where questions turn: In a conditional discrimination, a comparison is correct only in relation to the sample; correct choices that ignore the sample reflect a simple discrimination or faulty stimulus control.

Practise G.6

G.7Select and evaluate stimulus and response prompting procedures (e.g., errorless, most-to-least, least-to-most).

Response prompts act on the response: verbal instructions, modeling, and physical guidance (gestural prompts are often grouped with them); stimulus prompts act on the stimulus: within-stimulus (changing the size, color or position of the correct stimulus) and extra-stimulus (adding an extra cue, such as an arrow).
Most-to-least (MTL): start with the most intrusive prompt needed, then reduce intrusiveness; it reduces errors and suits new skills and learners who make many errors.
Least-to-most (LTM): give the learner the chance to respond independently and increase assistance step by step only when needed; it gives opportunities for independence but allows more errors.
Errorless learning: arrange prompts so the learner rarely makes errors, using MTL, time delay or stimulus fading; it reduces problem behavior linked to errors and prevents practicing incorrect responses.
Graduated guidance adjusts physical guidance moment by moment as the learner responds.
Select prompts by the learner's history, the skill, the risk of errors and the plan for fading; evaluate by data on prompted and independent responses and on prompt dependence.

In practice: A behavior analyst teaches a teenager to wash his hands with MTL prompting: full physical guidance at first, then partial guidance, then a model, then a gesture. With his peer, who already completes most steps, she uses LTM, giving a verbal prompt only when a step is missed.

Where questions turn: Most-to-least starts with the most help and reduces it; least-to-most starts with independence and adds help only when needed.

Practise G.7

G.8Design and implement procedures to fade stimulus and response prompts (e.g., prompt delay, stimulus fading).

Prompts must be faded so behavior comes under control of natural stimuli; prompt dependence results when prompts are faded too slowly or not at all.
Prompt delay (time delay): the prompt is delayed after the SD, either by a constant delay (0 seconds at first, then a fixed delay) or a progressive delay (the delay grows gradually), giving the learner a chance to respond before the prompt.
Stimulus fading: a stimulus prompt is gradually removed by changing a feature (fading a highlighted color, reducing the size of a cue) until only the natural stimulus remains.
Stimulus shaping: the configuration of the stimulus itself is gradually changed (a drawing of an object gradually changed into the written word) to transfer control.
MTL and LTM are also fading strategies; criteria for moving to a less intrusive prompt are set in advance (such as two consecutive correct sessions at a level).
Monitor errors as prompts fade: if errors increase, go back a step and fade more gradually.

In practice: A teacher says what is your name and immediately models the answer (0-second delay) for a few sessions, then waits 4 seconds before modeling. The child begins answering before the model and soon answers without it.

Where questions turn: In prompt delay, what changes is the time before the prompt is given, not the type of prompt; stimulus fading changes a feature of the stimulus prompt itself.

Practise G.8

G.9Design and evaluate modeling procedures.

Modeling is demonstrating a behavior so the learner imitates it; it can be live, symbolic (pictures, videos), video self-modeling (the learner sees an edited video of themselves performing correctly), or peer modeling.
Prerequisites: the learner must attend to the model and have a generalized imitation repertoire; teach imitation first if needed.
Effective models are similar to the learner, competent, and seen receiving reinforcement for the behavior; demonstrations are clear, of short steps, repeated and followed by an opportunity to imitate with feedback.
Modeling is part of behavioral skills training (BST): instructions, modeling, rehearsal and feedback, used for teaching clients, parents and staff.
Evaluate by whether the learner imitates the model accurately, whether imitation occurs in natural settings, and whether the skill is maintained without the model.

In practice: A behavior analyst creates a short video of a peer asking classmates to join a game. A 10-year-old watches it before recess and then practices with feedback; within a week he asks to join games at recess without watching the video first.

Where questions turn: Modeling only works if the learner has the prerequisite skills of attending and imitating; a learner without generalized imitation needs imitation training before modeling.

Practise G.9

G.10Design and evaluate instructions and rules.

Instructions and rules are verbal stimuli that describe what to do and often the contingency (if you do X, Y will happen); they produce rule-governed behavior that can change quickly without direct contact with consequences.
Effective instructions are clear, specific, brief, stated positively (what to do), given when the person attends, at their language level and followed by reinforcement for compliance.
Rules that specify the behavior, the context and the consequence (including delayed consequences) are more effective than vague statements.
Rule following depends on a history of reinforcement for following rules and on the credibility of the person giving them.
Rule-governed behavior can be insensitive to changes in contingencies; when rules no longer describe consequences accurately, behavior may persist or stop abruptly.
Evaluate by measuring compliance and adjusting the wording, timing or consequences; pair rules with consequences that make following them worthwhile.

In practice: Instead of telling a group be good during the field trip, the behavior analyst says stay with your buddy and walk beside the teacher; every 15 minutes that you do, you earn a sticker toward the zoo shop. Compliance improves compared with the vague instruction.

Where questions turn: A rule that does not describe the behavior specifically, or that is never followed by its stated consequence, is unlikely to control behavior over time.

Practise G.10

G.11Shape dimensions of behavior.

Shaping is the differential reinforcement of successive approximations to a terminal behavior: responses closer to the target are reinforced, and earlier approximations are put on extinction.
Shaping across topographies changes the form of the response (from making a sound to saying a word); shaping within a topography changes a dimension of the same response (rate, duration, latency, magnitude, accuracy).
Steps: define the terminal behavior, identify an initial behavior the person already emits that resembles it, choose the approximations and criteria for moving on, reinforce each approximation until stable, then move to the next.
The size of steps matters: too large produces extinction of responding, too small slows progress; adjust based on the learner's responses.
Shaping can produce undesired behavior unintentionally (gradually reinforcing louder tantrums), which is why reinforcing escalating behavior is avoided.
Shaping can be combined with prompting to speed it up, or with percentile schedules that define criteria by recent performance.

In practice: A child engages in a reading task for only 2 minutes. The teacher reinforces 3 minutes of reading, then 5 once 3 is reliable, then 8, building to 15 minutes (shaping duration within a topography).

Where questions turn: Shaping changes the response requirement for reinforcement across approximations; fading changes the antecedent prompt while the response stays the same.

Practise G.11

G.12Select and implement chaining procedures.

A behavior chain is a sequence of responses in which each response produces a stimulus change that reinforces it and serves as the SD for the next; the terminal reinforcer follows the last response.
A task analysis breaks the chain into teachable steps, developed by observing competent performers, consulting experts, or performing the task oneself; it is validated and individualized.
Forward chaining: teach the first step to criterion, then the first two, and so on, with the rest of the chain completed by the trainer or prompted; it suits learners who can learn early steps easily.
Backward chaining: complete all steps except the last for the learner, teach the last to criterion, then the last two, and so on; the learner contacts the terminal reinforcer on every trial. Backward chaining with leaps ahead skips steps the learner already masters.
Total-task chaining: the learner attempts every step in each trial, with prompts as needed; it suits learners who can do many steps already or when the chain is short.
Select by the learner's existing skills (measured by a single-opportunity or multiple-opportunity probe of the task analysis), the length and complexity of the chain, and the learner's motivation.

In practice: A behavior analyst teaches a young adult to make a sandwich with backward chaining: staff do every step but spreading the final layer and closing it, which she does herself and then eats. Each week she does one more step from the end until she completes the task independently.

Where questions turn: In backward chaining the learner completes the last step first and reaches the terminal reinforcer on every trial; in forward chaining the first step is taught first.

Practise G.12

G.13Design and evaluate trial-based and free-operant procedures.

Trial-based procedures, such as discrete-trial teaching, present instruction in distinct trials: an SD, a response, a consequence, and an intertrial interval; the instructor controls the pace and the number of opportunities.
Free-operant procedures let the response occur at any time with no restriction on opportunities (such as play-based learning, incidental teaching and naturalistic teaching), with learning arranged around the person's interests.
Discrete-trial teaching is efficient for many repetitions and clear stimulus control, but risks prompt dependence and limited generalization; naturalistic procedures promote generalization and spontaneous use and use the learner's motivation, but give fewer opportunities in the same time.
Naturalistic developmental behavioral interventions (for example, pivotal response treatment and incidental teaching) embed teaching in play and routines and use natural reinforcers.
Design: choose based on the skill, the learner and the goal; combine both when useful, with DTT to establish a skill and naturalistic practice to generalize it.
Evaluate by the rate of acquisition, generalization to natural settings, spontaneous use and the learner's engagement and assent.

In practice: A child learns color names quickly in discrete trials at a table, but does not use them at home. The behavior analyst adds incidental teaching during play, waiting for the child to want a colored block and prompting him to ask for the red one.

Where questions turn: Free-operant means the learner can respond at any time with unrestricted opportunities; in trial-based teaching the instructor controls when each opportunity occurs.

Practise G.13

G.14Design and evaluate group contingencies.

Independent group contingency: the same contingency applies to every member, but each person's reinforcement depends only on their own behavior (any student who finishes the work earns free time).
Dependent group contingency: the reinforcement for the whole group depends on the behavior of one individual or a small group (the class earns a party if a specific student meets his goal).
Interdependent group contingency: all members must meet the criterion, or the group as a whole must meet it, for anyone to receive reinforcement (the class earns extra recess if the class average meets a goal, or the Good Behavior Game with teams).
Advantages: efficiency, use of peer influence, and promotion of cooperation; risks: peer pressure or harassment of a member who fails, some members sabotaging, and some members not contributing.
Design: define behaviors clearly, set achievable criteria, choose reinforcers valued by the group, monitor for negative peer interactions, and consider combining with individual contingencies.
Evaluate by the group's and individuals' performance and by checking for unwanted social side effects.

In practice: In the Good Behavior Game, a class is divided into teams, and each team that has fewer than 5 marks for rule violations during the lesson earns a reward. All members of a team share the outcome, an interdependent contingency.

Where questions turn: In a dependent contingency the group's reward rests on one person's or a few persons' behavior; in an interdependent contingency it rests on everyone or on the group as a whole.

Practise G.14

G.15Design and evaluate procedures to promote stimulus and response generalization.

Generalization must be planned, not hoped for (train and hope is the weakest strategy).
Teach enough examples: train with multiple exemplars of stimuli and responses; general case analysis selects examples that represent the full range of variations in the natural setting.
Make the training setting similar to the generalization setting: program common stimuli (bring elements of the natural setting into teaching, or vice versa) and train loosely (vary non-critical aspects of instruction, such as instructor, wording and location).
Use indiscriminable contingencies (intermittent or delayed reinforcement so the learner cannot tell when reinforcement is available) and recruit natural contingencies, such as teaching the learner to recruit reinforcement or using behavior traps.
Mediate generalization (teach self-management or a verbal mediator the learner carries across settings) and train to generalize (reinforce generalized responses).
Evaluate with probes in untrained settings, with untrained people and stimuli, and for untrained responses.

In practice: To teach purchasing, a behavior analyst teaches with several stores, cashiers, items and payment types selected to represent the variations the learner will meet, then probes in a new store. The learner buys items successfully in the new store.

Where questions turn: Teaching one perfect example in one setting with one instructor produces tight stimulus control, which often prevents generalization.

Practise G.15

G.16Design and evaluate procedures to maintain desired behavior change following intervention (e.g., schedule thinning, transferring to naturally occurring reinforcers).

Maintenance is planned from the start, by choosing behaviors that will contact natural reinforcement and by teaching to fluency.
Schedule thinning: move from continuous to intermittent reinforcement gradually, increasing the ratio or interval to levels comparable to the natural environment.
Transfer to naturally occurring reinforcers: pair contrived reinforcers with natural ones (praise with tokens), fade contrived reinforcers, and teach behaviors that produce natural reinforcement (communication that gets real results).
Delay reinforcement gradually, teach self-management (self-monitoring, self-recording, self-reinforcement), and involve caregivers and natural change agents.
Booster sessions and periodic probes detect and address decline.
Evaluate by measuring the behavior after the intervention is withdrawn and over time.

In practice: A teenager learns to greet coworkers with tokens for each greeting. The behavior analyst moves to tokens for every third greeting, then only at the end of the day, while coworkers greet him warmly in return; six months later he still greets them without tokens.

Where questions turn: Abruptly removing reinforcement after acquisition produces extinction, not maintenance; reinforcement is thinned gradually and replaced by natural reinforcement.

Practise G.16

G.17Design and evaluate positive and negative punishment (e.g., time-out, response cost, overcorrection).

Positive punishment procedures add a stimulus contingent on behavior: reprimands, response blocking (which may function as punishment or extinction), contingent exercise and overcorrection (restitutional overcorrection restores the environment to a better state than before; positive practice overcorrection requires repeated practice of the correct behavior).
Negative punishment procedures remove a stimulus: time-out from positive reinforcement (nonexclusion types such as planned ignoring, withdrawal of a specific reinforcer, contingent observation and the time-out ribbon; exclusion time-out removes the person from the setting) and response cost (loss of a specific amount of a reinforcer, as a fine or a bonus response cost).
Time-out works only if the time-in environment is reinforcing and if time-out does not allow escape from demands; with escape-maintained behavior, time-out may reinforce the behavior.
Design: use punishment only after reinforcement-based and less intrusive procedures have been considered or tried, with informed consent, clear definitions, immediate and consistent delivery, and a reinforcement procedure for alternative behavior.
Evaluate with data from the start: a punishment procedure that does not reduce behavior quickly should be stopped; monitor side effects and fade the procedure.
Follow organizational, legal and ethical requirements on restrictive procedures, including review and supervision.

In practice: A student throws materials during independent work. The team first tries reinforcement of work and teaching requests for help; when throwing persists, with parental consent they add a response cost of one token per throw while continuing to reinforce work, and throwing drops within a week.

Where questions turn: Time-out for behavior maintained by escape can function as negative reinforcement, because it removes the person from the demand.

Practise G.17

G.18Evaluate emotional and elicited effects of behavior-change procedures.

Extinction can produce an extinction burst, emotional responding (crying, anger) and extinction-induced aggression, as well as resurgence and spontaneous recovery.
Punishment can elicit emotional responses and aggression, produce escape and avoidance of the setting or the person delivering it, and model the use of aggression; the person delivering punishment can be negatively reinforced for using it, which can lead to overuse.
Respondent conditioning can occur during procedures: stimuli present during aversive events (a room, a person, a task) can come to elicit fear or emotional responses.
Reinforcement procedures have fewer emotional side effects but can produce problems: satiation, reliance on contrived reinforcers, and emotional responses when reinforcement is thinned too quickly (ratio strain).
Evaluate by measuring emotional and elicited responses (crying, aggression, avoidance) as well as the target behavior, and by asking about the client's assent and comfort.
Reduce side effects by combining procedures with reinforcement of alternatives, enriching the environment, introducing extinction or punishment only with planned safety measures, and monitoring closely.

In practice: When escape extinction is introduced for a child's refusal to eat, he cries and attempts to leave the table more than before. The behavior analyst measures these responses, adds reinforcement for each bite and a choice of foods, and the emotional responses decrease within a week.

Where questions turn: Emotional responses during extinction are elicited side effects that should be anticipated and measured, not signs that the procedure is working or failing by themselves.

Practise G.18

G.19Design and evaluate procedures to promote emergent relations and generative performance.

Procedures that promote emergent relations train some relations so that others emerge without training, increasing the efficiency of teaching.
Equivalence-based instruction: train conditional discriminations (A to B, B to C) with matching-to-sample in a training structure (one-to-many, many-to-one or linear series), then test untrained relations (B to A, A to C, C to A) without reinforcement.
Matrix training: teach a few combinations of elements arranged in a matrix (such as actions and objects), selected so that each element is taught in some combination, then test untrained combinations (recombinative generalization).
Multiple exemplar instruction across operants (teaching the same stimulus as tact, listener response and intraverbal) can promote bidirectional relations such as naming.
Instructive feedback presents extra, non-target information during consequences, which learners often acquire without direct teaching.
Evaluate by testing for emergent responses in probe trials without reinforcement, and by measuring how many responses emerged relative to those trained.

In practice: A behavior analyst teaches a learner to match the spoken names of 5 coins to the coins and the coins to their printed values. She then tests, without reinforcement, whether the learner matches printed values to coin names; the learner does so for 4 of 5 coins.

Where questions turn: Testing for emergent relations is done without reinforcement for the tested relations; reinforcing them during the test turns the test into training.

Practise G.19

Key terms

DRA: Differential reinforcement of an alternative behavior while withholding reinforcement for problem behavior.
DRO: Reinforcement delivered when problem behavior has not occurred for an interval or at a moment.
DRL: Reinforcement of responding at a low rate or after a minimum interresponse time.
Noncontingent reinforcement: Delivery of the maintaining reinforcer on a time-based schedule independent of behavior.
Token economy: A system of conditioned generalized reinforcers exchangeable for backup reinforcers.
Most-to-least prompting: Starting with the most intrusive prompt and reducing assistance.
Least-to-most prompting: Starting with an independent opportunity and increasing assistance only when needed.
Prompt delay: Fading by increasing the time between the SD and the prompt.
Shaping: Differential reinforcement of successive approximations to a terminal behavior.
Backward chaining: Teaching a chain from the last step, with the learner contacting the terminal reinforcer each trial.
Interdependent group contingency: Reinforcement for all depends on all members or the group as a whole meeting the criterion.
General case analysis: Selecting teaching examples that represent the range of variations in the natural environment.
Overcorrection: Restitution or positive practice required contingent on problem behavior.
Exclusion time-out: Removing the person from the setting for a period contingent on behavior.
Matrix training: Teaching some combinations of elements so untrained combinations emerge.

After this guide: Given a case, you should be able to design, select or evaluate the behavior-change procedure that fits the function and the goal, and say what to change when the data show it is not working or producing side effects.

Practise area G