Study guide · Area B

Concepts and Principles

24 of the exam's 175 scored questions (14%) · 24 tasks

This area covers the basic principles every procedure rests on: behavior and stimuli, respondent and operant conditioning, reinforcement and punishment, schedules, extinction, stimulus control, motivating operations, verbal behavior, emergent relations, momentum, matching and imitation. Questions describe a short episode and ask you to classify it, so the skill being tested is reading the effect on future behavior and the exact operation, not the intention or the label someone gave it.

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 (24 of them), in the BACB's words.

B.1Identify and distinguish among behavior, response, and response class.

Behavior is the activity of living organisms: the portion of an organism's interaction with its environment that involves movement of some part of the organism. The dead man test applies: if a dead man can do it, it is not behavior.
A response is a single instance of behavior, one occurrence that can be observed and measured (one hand raise, one word spoken).
A response class is a group of responses that share an effect on the environment (a functional response class) or a common form (a topographical response class); in applied work the functional sense matters most.
Members of a functional response class can look very different: screaming, hitting and dropping to the floor may all produce escape from tasks and so belong to one class, and reinforcing or extinguishing one member tends to affect the others.
A repertoire is all the behavior a person can do, or the set of behaviors relevant to a skill area.

In practice: A child who wants to stop a puzzle pushes it away, says no and hits the table; each single push, each no and each hit is one response. Because all three end the task, they form one functional response class, and the analyst plans to teach a request for a break that serves the same function rather than targeting each response separately.

Where questions turn: Responses that look nothing alike can be one response class if they produce the same effect, and responses that look alike can belong to different classes if they serve different functions.

Practise B.1

B.2Identify and distinguish between stimulus and stimulus class.

A stimulus is an energy change that affects an organism through its receptor cells; stimuli can be described formally (physical features), temporally (before or after a behavior) and functionally (their effect on behavior).
A stimulus class is a group of stimuli that share a common element: formal (same physical features, such as all red objects), temporal (all occurring before a behavior, antecedents, or after it, consequences) or functional (all producing the same effect on behavior).
Members of a functional stimulus class can differ in form: the printed word dog, a photograph of a dog and the spoken word dog can all evoke the same response.
The environment is the full set of circumstances in which the organism exists; for analysis it is broken into specific stimuli and stimulus classes that change behavior.
A stimulus class matters because responding trained to one member tends to occur to other members; this is the basis for generalization and for arbitrary classes such as stimulus equivalence.

In practice: A learner who has been taught to say cup to her own blue cup also says cup to a paper cup, a mug and a picture of a glass of water. The analyst describes these as members of a stimulus class that now evoke the same tact.

Where questions turn: One stimulus is a single energy change; a stimulus class is a set of stimuli grouped by shared features, timing or function, and an exam item will describe several different objects controlling one response to test whether you see a class.

Practise B.2

B.3Identify and distinguish between respondent and operant conditioning.

Respondent behavior is elicited by antecedent stimuli and is controlled by them; the basic unit is the reflex, an unconditioned stimulus (US) that elicits an unconditioned response (UR) without learning.
Respondent conditioning is stimulus-stimulus pairing: a neutral stimulus is presented with or just before a US until it alone elicits a similar response; it is then a conditioned stimulus (CS) eliciting a conditioned response (CR).
Higher-order conditioning occurs when a neutral stimulus is paired with an existing CS rather than a US and comes to elicit the response.
Operant behavior is behavior whose future frequency is determined by its history of consequences; it is emitted, and the stimuli before it set the occasion for it rather than eliciting it.
Operant conditioning is a stimulus-response-consequence relation (the three-term contingency, with motivating operations often added as a fourth term); responses are selected by the consequences that follow them.
The quick test: if the behavior is a reflexive bodily response produced by a preceding stimulus (salivation, flinching, heart rate), think respondent; if it changes because of what happened after it in the past, think operant. Most real episodes include both.

In practice: After several painful injections, a child's heart races and he cries when he sees the nurse's white coat: the coat has become a CS eliciting a CR. When he then runs out of the room and the injection is postponed, running out is operant behavior strengthened by escape, and it will occur more in future.

Where questions turn: Respondent conditioning pairs two stimuli and does not depend on what the person does; operant conditioning depends on a consequence of a response, so a pairing of stimuli with no response requirement is never operant.

Practise B.3

B.4Identify and distinguish between positive and negative reinforcement contingencies.

Reinforcement is a response followed immediately by a change in stimulus that increases the future frequency of similar responses; it is defined by its effect, not by whether the consequence seems pleasant.
Positive reinforcement: a stimulus is presented (added) after the response and the response increases in future. Negative reinforcement: a stimulus is removed or reduced after the response and the response increases in future.
Positive and negative refer to adding or removing a stimulus, never to good or bad; both forms strengthen behavior.
Escape is negative reinforcement in which the response ends a stimulus already present; avoidance is negative reinforcement in which the response prevents or postpones a stimulus before it occurs (discriminated avoidance has a warning signal, free-operant avoidance does not).
To classify, ask three questions in order: did the behavior increase in future, was a stimulus added or taken away, and what stimulus was it.

In practice: Each time a teenager puts on his seatbelt, the car's warning chime stops, and he now buckles up faster each trip. The response removes an ongoing aversive stimulus and increases, so it is negative reinforcement (escape); if he buckles before the chime ever starts, it is avoidance.

Where questions turn: Negative reinforcement is not punishment; it increases behavior by removing a stimulus, and a question that describes something unpleasant being taken away after a response that then increases is negative reinforcement.

Practise B.4

B.5Identify and distinguish between positive and negative punishment contingencies.

Punishment is a response followed immediately by a change in stimulus that decreases the future frequency of similar responses; like reinforcement, it is defined only by its effect on future behavior.
Positive punishment: a stimulus is presented after the response and the response decreases in future (a reprimand, a required effortful task). Negative punishment: a stimulus is removed after the response and the response decreases in future (loss of tokens, loss of access to an activity).
Negative punishment comes in two procedural forms: time-out from positive reinforcement (loss of access to reinforcers for a period) and response cost (loss of a specific amount of a reinforcer already held).
If a consequence intended as punishment does not decrease the behavior, it is not punishment; a scolding that is followed by more of the behavior may be functioning as positive reinforcement.
Recovery from punishment occurs when the punishment contingency is discontinued and the behavior returns toward its earlier level.

In practice: A child loses five minutes of tablet time each time she throws food, and food throwing drops over the following week. A stimulus was removed and the behavior decreased in future, so the contingency is negative punishment (response cost, if the minutes were a specific amount already earned).

Where questions turn: A procedure is called punishment only when the behavior decreases; an aversive event that the person dislikes but that does not reduce the behavior is not punishment in the technical sense.

Practise B.5

B.6Identify and distinguish between automatic and socially mediated contingencies.

A socially mediated contingency is one in which another person delivers the consequence: attention, access to items, or the removal of a demand by a teacher or parent.
An automatic contingency is one in which the behavior produces its consequence directly, without anyone else's action: the sensory result of rocking, the relief of scratching an itch, the sight of lights flicked on and off.
Both kinds can be positive or negative: automatic positive reinforcement adds a sensory stimulus; automatic negative reinforcement removes one (scratching that reduces itching, pressing on a painful ear).
Automatic reinforcement is suggested when a behavior persists when the person is alone or across all conditions of a functional analysis, but the label describes the source of the consequence, not the absence of a known cause.
The distinction drives intervention: socially mediated behavior is treated by changing how others respond, while automatically maintained behavior needs strategies such as competing sensory stimulation, response blocking or environmental enrichment.

In practice: A man flaps paper near his eyes for long periods when alone in his room and in every condition of an analysis, regardless of whether anyone responds. The analyst hypothesizes automatic positive reinforcement and plans access to matched visual stimulation instead of planned ignoring, which would not remove the reinforcer.

Where questions turn: Automatic does not mean unlearned or without a consequence: the behavior is still operant and reinforced, just by its own sensory product rather than by another person.

Practise B.6

B.7Identify and distinguish among unconditioned, conditioned, and generalized reinforcers.

An unconditioned reinforcer (primary) functions as a reinforcer without any learning history because of the species' evolutionary history: food, water, warmth, sexual stimulation, and escape from painful stimulation.
A conditioned reinforcer (secondary) is a previously neutral stimulus that acquires its reinforcing function through pairing with one or more existing reinforcers: praise, a teacher's smile, a sticker.
A generalized conditioned reinforcer has been paired with many other reinforcers, so its effectiveness depends little on any one motivating operation: money and tokens exchangeable for many backup items are the standard examples.
The effectiveness of unconditioned reinforcers is strongly tied to deprivation and satiation; generalized reinforcers stay effective across states because some backup reinforcer is usually valued.
Conditioned reinforcers can lose their function if pairing with backup reinforcers stops, which is why token systems keep exchange opportunities available.

In practice: A teacher praises a student each time she hands him a token, and tokens buy recess time, snacks or computer time. Over weeks, praise alone starts to maintain on-task behavior (a conditioned reinforcer), and the tokens remain effective even after lunch because they can buy many things (a generalized conditioned reinforcer).

Where questions turn: A stimulus is not a generalized reinforcer simply because it is effective for many people; it is generalized because it has been paired with many different reinforcers for this person.

Practise B.7

B.8Identify and distinguish among unconditioned, conditioned, and generalized punishers.

An unconditioned punisher functions as a punisher without prior learning: painful stimulation, extreme temperatures, loud noises, intense light, and other stimuli the species evolved to be sensitive to.
A conditioned punisher is a previously neutral stimulus that acquires its punishing function through pairing with an existing punisher or with the loss of reinforcers: a frown, the word no, a warning signal.
A generalized conditioned punisher has been paired with many punishers or with the loss of many reinforcers, so it works across motivational states: disapproval from a parent, a fine.
Like reinforcers, punishers are defined by their effect: a stimulus that decreases one person's behavior may have no effect, or a reinforcing effect, on another's.
Conditioned punishers that are no longer paired with other punishers can lose their effect, and the person delivering punishment can itself become a conditioned aversive stimulus.

In practice: A toddler touches a hot stove and immediately pulls back; touching stoves decreases, an unconditioned punisher. His mother says hot sharply each time, and later the word alone stops him from reaching for a hot pan: a conditioned punisher.

Where questions turn: Conditioned and generalized punishers are acquired through pairing; an exam item that describes a word or look acquiring its effect through a history of association is describing a conditioned punisher, not an unconditioned one.

Practise B.8

B.9Identify and distinguish among simple schedules of reinforcement.

Continuous reinforcement (CRF) reinforces every response and is best for establishing new behavior; intermittent schedules reinforce only some responses and are best for maintaining it, producing greater resistance to extinction.
Fixed ratio (FR): reinforcement after a fixed number of responses; typically a high rate with a postreinforcement pause that lengthens as the ratio grows. Variable ratio (VR): reinforcement after a variable number of responses around an average; a high, steady rate with little or no pause.
Fixed interval (FI): reinforcement for the first response after a fixed time has elapsed; a postreinforcement pause followed by an accelerating rate (the FI scallop). Variable interval (VI): reinforcement for the first response after variable times around an average; a moderate, steady rate.
Ratio schedules generally produce higher rates than interval schedules because a faster rate yields reinforcement sooner; on interval schedules, responding faster does not bring reinforcement before the interval ends.
A limited hold adds a deadline: on an interval schedule, the reinforcer stays available only for a set time after the interval ends, which raises the rate of responding.
Ratio strain is the disruption of responding (pausing, avoidance, aggression) when a ratio is thinned too quickly; the remedy is to return to a smaller ratio and thin more gradually.

In practice: A worker paid for every 10 envelopes stuffed (FR 10) works quickly and pauses briefly after each batch. A student whose teacher checks her work at unpredictable times averaging every 5 minutes (VI 5 min) works at a steady moderate pace throughout the period.

Where questions turn: On an interval schedule the passage of time alone does not deliver the reinforcer; a response is still required after the interval ends, which separates FI and VI from fixed-time and variable-time schedules.

Practise B.9

B.10Identify and distinguish among concurrent, multiple, mixed, and chained schedules of reinforcement.

Concurrent schedules: two or more schedules are in effect at the same time, each for a different response, and the person chooses among them; they are the basis for studying choice and the matching law.
Multiple schedule: two or more basic schedules alternate, usually at random, each signaled by its own discriminative stimulus; each component has its own schedule (one may be extinction).
Mixed schedule: the same alternation of two or more schedules as a multiple schedule, but with no signal correlated with each component.
Chained schedule: two or more schedules occur in a fixed sequence, each signaled by its own stimulus; completing one component produces the stimulus for the next, and the terminal reinforcer comes only after the last. Tandem is the same arrangement without signals.
Related compound schedules: in an alternative schedule reinforcement follows whichever of two requirements is met first; in a conjunctive schedule both must be met.
To classify, ask: are the components available at the same time (concurrent) or in succession; if in succession, does each component end in reinforcement on its own (multiple or mixed) or does only completing the whole sequence produce it (chained or tandem); and is each signaled (multiple, chained) or not (mixed, tandem).

In practice: During a lesson, a card on the desk shows green when hand raising earns praise on an FR 1 and red when praise is available only on an extinction schedule; the two alternate through the morning. Because the components alternate and each is signaled, this is a multiple schedule, often used to thin functional communication by teaching when requests will be honored.

Where questions turn: Multiple and mixed schedules differ only in whether each component is signaled; chained and tandem differ the same way, and the difference between those two pairs is whether each component is reinforced or only the last one.

Practise B.10

B.11Identify and distinguish between operant and respondent extinction as operations and processes.

Operant extinction as an operation: reinforcement for a previously reinforced behavior is discontinued, so the response no longer produces its maintaining consequence.
Operant extinction as a process: the resulting decrease in the frequency of the behavior, often gradual and often preceded by an extinction burst (an immediate increase in frequency, duration, magnitude or variability).
Other extinction effects include spontaneous recovery (the behavior reappears after a period without it, even though extinction is still in place), emotional responding and extinction-induced aggression, and resurgence of previously reinforced behavior.
Respondent extinction as an operation: the conditioned stimulus is presented repeatedly without the unconditioned stimulus. As a process: the conditioned stimulus gradually loses its ability to elicit the conditioned response.
Extinction must match function: for attention-maintained behavior, attention is withheld; for escape-maintained behavior, the demand stays in place (escape extinction); for automatically maintained behavior, the sensory consequence is masked or blocked (sensory extinction).
Extinction is not punishment: nothing is added or removed after the response, the maintaining reinforcer simply no longer follows it.

In practice: A child who used to get a cookie for whining in the store now gets none; whining first grows louder and longer for two trips (an extinction burst), then drops, then briefly returns a month later (spontaneous recovery). Separately, a child afraid of dogs after a bite who meets many calm dogs with no further bites shows respondent extinction as fear responses weaken.

Where questions turn: Removing a stimulus the person already had is negative punishment; withholding the specific reinforcer that used to follow the response is extinction.

Practise B.11

B.12Identify examples of stimulus control.

Stimulus control exists when the rate, latency, duration or magnitude of a response is altered in the presence of an antecedent stimulus; the behavior occurs more often (or differently) when the stimulus is present than when it is absent.
Stimulus control develops through differential reinforcement: a response is reinforced in the presence of one stimulus and not, or less, in its absence.
A discriminative stimulus (SD) is a stimulus in whose presence a response has been reinforced and which therefore evokes the response; an S-delta is a stimulus in whose presence the response has not been reinforced.
A discriminative stimulus for punishment (SDp) is a stimulus in whose presence a response has been punished, which suppresses the response.
Stimulus control is the goal of most instruction: a skill counts as learned only when it occurs to the right stimuli (answering what is your name, stopping at red lights).

In practice: Children on a playground run and shout freely but walk and whisper as soon as the principal appears, because quiet behavior has been reinforced and loud behavior reprimanded in her presence. The principal's presence is exerting stimulus control over the children's behavior.

Where questions turn: A stimulus exerts control because of a history of differential consequences in its presence, not because it forces or elicits the behavior; an operant SD evokes behavior, a respondent CS elicits it.

Practise B.12

B.13Identify examples of stimulus discrimination.

Stimulus discrimination is responding differently in the presence of different stimuli: the response occurs to the SD and not, or less, to the S-delta.
Discrimination is produced by discrimination training: reinforcing a response in the presence of one stimulus and extinguishing it in the presence of others.
Discrimination and generalization lie on a continuum: a steep stimulus generalization gradient shows tight discrimination, a flat one shows broad generalization.
Faulty stimulus control is a discrimination based on an irrelevant feature, such as a learner who picks the correct card because it is always on the left rather than because of its picture.
Discrimination is often a teaching goal in itself: telling a stranger from a family member, a five-dollar bill from a one-dollar bill.

In practice: A child raises his hand to ask questions in his classroom, where teachers answer, but not in the library, where staff have never responded to a raised hand. His behavior shows stimulus discrimination between the two settings.

Where questions turn: A learner who answers correctly only when the correct choice is in the same position is showing discrimination, but to the wrong stimulus; correct answers alone do not prove the intended stimulus control.

Practise B.13

B.14Identify and distinguish between stimulus and response generalization.

Stimulus generalization: a response reinforced in the presence of one stimulus also occurs in the presence of other, untrained stimuli that share physical properties with it.
Response generalization: training one response produces untrained responses that are functionally equivalent to it, different forms that produce the same outcome.
Put simply: stimulus generalization is the same response to new stimuli; response generalization is new responses to the same stimulus or for the same function.
Overgeneralization is a problem form of stimulus generalization, as when a child calls every man daddy.
Generalization is an outcome that applied programs plan for and measure; it is not assumed to happen on its own.

In practice: After being taught to greet her teacher with hello, a girl also says hello to the bus driver and the cafeteria staff (stimulus generalization). She also begins greeting her teacher with hi and good morning, which were never taught (response generalization).

Where questions turn: Ask which part is new: if the antecedent situation is new and the response is the same, it is stimulus generalization; if the response form is new, it is response generalization.

Practise B.14

B.15Identify examples of response maintenance.

Response maintenance is the extent to which a learner continues to perform a target behavior after part or all of the intervention that produced it has been withdrawn.
Maintenance is measured over time: probes days, weeks or months after training ends show whether the behavior persists under the natural conditions.
Maintenance differs from generalization across settings or behaviors: maintenance concerns persistence over time in the absence of the full intervention.
Behavior is more likely to be maintained when it contacts naturally existing contingencies of reinforcement, when reinforcement has been thinned to an intermittent schedule, and when the learner can manage some contingencies themselves.
A behavior that returns to baseline levels as soon as the program stops has not been maintained, however well it was acquired.

In practice: A man learns to pack his lunch with a picture schedule and praise. Two months after the schedule and the praise are removed, he still packs a complete lunch each workday because his coworkers comment on his meals; that continued performance is response maintenance.

Where questions turn: Performance during the intervention, however high, is acquisition; maintenance is shown only after the intervention has been partly or fully withdrawn.

Practise B.15

B.16Identify examples of motivating operations.

A motivating operation (MO) is an environmental variable that alters the effectiveness of some stimulus as a reinforcer or punisher (the value-altering effect) and alters the current frequency of all behavior that has been reinforced or punished by that stimulus (the behavior-altering effect).
An establishing operation (EO) increases the reinforcing effectiveness of a stimulus and evokes behavior that has produced it; an abolishing operation (AO) decreases its effectiveness and abates that behavior.
Unconditioned MOs (UMOs) have value-altering effects without learning: deprivation and satiation of food, water, sleep, activity; temperature changes; painful stimulation.
Conditioned MOs (CMOs) work because of a learning history. Surrogate (CMO-S): a stimulus paired with another MO takes on its effect. Reflexive (CMO-R): a stimulus that precedes a worsening makes its own removal a reinforcer, as when a demand signals work (one that precedes an improvement makes its own removal a punisher). Transitive (CMO-T): a stimulus makes another stimulus effective as a reinforcer, as when a locked door makes the key valuable.
Behavior-altering effects are evocative (increase current frequency) or abative (decrease current frequency); their effect on behavior is momentary, unlike the lasting function-altering effect of a reinforcer.

In practice: A child who skipped his afternoon snack asks for food repeatedly at four o'clock, and the snack is especially powerful as a reinforcer for finishing homework (food deprivation, a UMO acting as an EO). Later, when his parent hands him a toy that needs batteries, asking for batteries suddenly increases: the toy is a CMO-T that makes batteries valuable.

Where questions turn: An MO changes how much a consequence is worth right now and how often behavior that produced it occurs; it does not signal that the consequence is available, which is the job of an SD.

Practise B.16

B.17Distinguish between motivating operations and stimulus control.

Both MOs and SDs are antecedents that evoke behavior, which is why they are confused; they differ in why they evoke it.
An SD evokes behavior because the reinforcer has been more available in its presence than in its absence (differential availability).
An MO evokes behavior because it changes the current effectiveness of the reinforcer, whether or not the reinforcer is available.
A test: ask whether the behavior would have been reinforced in the absence of the antecedent. If the reinforcer would have been just as available, but no longer valuable, the antecedent is an MO; if the reinforcer would have been unavailable, it is an SD.
The CMO-T is the classic case: the screw that needs a screwdriver is an MO for asking for the screwdriver (the screwdriver's value goes up); the presence of a person who has screwdrivers is the SD for asking.

In practice: A child asks for water only after running outside in the heat, and only when an adult is nearby. Being hot and thirsty makes water valuable (an MO); the adult's presence signals that asking will produce water (an SD), since asking when alone has never worked.

Where questions turn: If the antecedent changes how much the person wants the consequence, it is an MO; if it changes the likelihood that the consequence will follow the response, it is an SD.

Practise B.17

B.18Identify and distinguish between rule-governed and contingency-shaped behavior.

Contingency-shaped behavior is behavior selected by direct contact with its consequences, through a history of reinforcement and punishment.
Rule-governed behavior is behavior controlled by a rule, a verbal description of a contingency (antecedent, behavior, consequence), rather than by direct contact with that contingency.
Rules allow behavior to come under control of consequences that are too delayed, too small, too improbable or too cumulative to shape behavior directly, such as saving for retirement or wearing a seatbelt.
Rule-governed behavior can be insensitive to changes in the actual contingencies: a person may keep following a rule even when it no longer describes what happens.
Rule following is itself operant behavior, developed through a history of reinforcement for following rules (and the source of the rule matters).

In practice: A new employee arrives on time because her manager told her that three late arrivals lead to a warning, although she has never received one: her punctuality is rule-governed. Her coworker, never told any policy, now avoids the highway ramp because, over many mornings, turning onto it was followed at once by standstill traffic while the side street kept her moving; she cannot state a rule for it: her route choice is contingency-shaped.

Where questions turn: Behavior that changes when a consequence is delayed by days or weeks is almost certainly rule-governed, because consequences that remote do not directly reinforce or punish behavior.

Practise B.18

B.19Identify and distinguish among verbal operants.

Mand: verbal behavior evoked by an MO and reinforced by the specific consequence specified (asking for juice when thirsty, reinforced by juice). It benefits the speaker directly.
Tact: verbal behavior evoked by a nonverbal stimulus (an object, event or property) and reinforced by generalized conditioned reinforcement, such as praise (saying dog on seeing a dog).
Echoic: verbal behavior evoked by a verbal stimulus, with point-to-point correspondence and formal similarity (same sense mode and physical resemblance) between stimulus and response (saying ball after hearing ball).
Intraverbal: verbal behavior evoked by a verbal stimulus with no point-to-point correspondence (answering four to what is two plus two, saying blue to the sky is).
Textual: reading aloud, a vocal response evoked by a written stimulus, with point-to-point correspondence but no formal similarity. Transcription: writing or spelling a spoken word, with point-to-point correspondence but no formal similarity. Copying a text: writing what is seen, with both.
Autoclitic: secondary verbal behavior that depends on and modifies the speaker's own verbal behavior (I think, not, the order of words), changing its effect on the listener.

In practice: A learner sees a cookie and says cookie when her teacher asks what is this (a tact, if praise follows), says cookie when she is hungry and no cookie is visible (a mand, reinforced by the cookie), and says cookie after hearing the teacher say cookie (an echoic).

Where questions turn: The same word can be any operant; classify it by the controlling antecedent and the consequence, not by the form of what was said.

Practise B.19

B.20Identify the role of multiple control in verbal behavior.

Most verbal responses are controlled by more than one variable at once (Skinner's multiple causation); a full analysis identifies all of them.
Convergent multiple control: a single verbal response is controlled by more than one variable at the same time, such as a mand and a tact together (saying ball when both seeing a ball and wanting it).
Divergent multiple control: a single variable affects the strength of many different responses, as when the question what animals live on farms evokes cow, pig, horse and goat.
Many intraverbals combine several controlling variables: the verbal stimulus, the setting, a visual stimulus and motivation may all help evoke one answer.
Multiple control explains verbal phenomena such as puns, slips of the tongue and creative combinations, and it guides teaching: a response can be strengthened under one source of control and then transferred to another.

In practice: A child names a red truck in a picture when the teacher asks what color is it: the picture (tact control) and the question (intraverbal control) both contribute to red, which is convergent control. Asked to name things that are red, she says apple, fire truck and stop sign, which is divergent control.

Where questions turn: Convergent control is many variables to one response; divergent control is one variable to many responses.

Practise B.20

B.21Identify examples of processes that promote emergent relations and generative performance.

An emergent relation is a stimulus relation or response that appears without having been directly trained or reinforced, derived from other relations that were trained.
Stimulus equivalence is demonstrated by three untrained relations after training (for example A to B and B to C): reflexivity (A to A, identity matching), symmetry (B to A) and transitivity (A to C); the combined test of symmetry and transitivity (C to A) is often called the equivalence test.
Relational frame theory describes derived relations beyond sameness (opposition, comparison, hierarchy, time) through mutual entailment, combinatorial entailment and the transformation of stimulus functions.
Generative performance also includes recombinative generalization, in which learned units combine into untrained combinations (taught red ball and blue car, the learner says red car), and generalized imitation and generalized operants such as identity matching.
These processes make teaching efficient: a few trained relations can produce many more untrained ones, which is why curricula are arranged to promote them.

In practice: A learner is taught to match the spoken word apple to a picture of an apple, and the picture to the printed word apple. Without further training, she now matches the printed word to the spoken word and reads it aloud: emergent relations from stimulus equivalence.

Where questions turn: A relation that was directly reinforced is not emergent, however advanced it looks; the defining feature is that it appears without direct training.

Practise B.21

B.22Identify ways behavioral momentum can be used to understand response persistence.

Behavioral momentum is a metaphor (Nevin) for the persistence of behavior: just as an object's momentum depends on its mass and velocity, behavior's resistance to change depends on its history of reinforcement.
Resistance to change is how much a behavior persists when a disruptor is applied, such as extinction, satiation, distraction or an alternative source of reinforcement.
In behavioral momentum theory, resistance to change depends on the rate of reinforcement obtained in a stimulus context (the stimulus-reinforcer relation), while response rate depends more on the response-reinforcer relation.
A practical implication: adding noncontingent or alternative reinforcement in the same context can reduce the rate of problem behavior while increasing its persistence when the alternative reinforcement stops; teaching the alternative in a different context and then combining them is one tactic studied for this reason.
The high-probability request sequence uses the momentum idea: several easy requests with a high likelihood of compliance, each reinforced, are given in quick succession before a low-probability request.

In practice: Before asking a reluctant student to start writing, a teacher gives three quick requests he usually follows (give me five, touch your nose, pass the pencil), praising each, and then asks him to write his name; he complies. The teacher is using high-probability requests to build compliance that persists into the harder request.

Where questions turn: Momentum theory says persistence is governed by the reinforcement in the context as a whole, so more reinforcement for an alternative in that context can make the target behavior more persistent, not less.

Practise B.22

B.23Identify ways the matching law can be used to interpret response allocation.

The matching law (Herrnstein) states that, on concurrent schedules, the relative rate of responding to an alternative tends to match the relative rate of reinforcement obtained from it.
Response allocation also depends on the quality, magnitude, immediacy of reinforcement and on the effort each response requires, not only on its rate.
The generalized matching law adds sensitivity (undermatching, responding less extreme than reinforcement ratios predict, or overmatching, more extreme) and bias (a consistent preference for one alternative not explained by reinforcement).
Problem behavior is a choice among concurrently available responses; it can be reduced by making the alternative response produce richer, faster, better reinforcement with less effort, and by reducing reinforcement for the problem behavior.
The matching law helps explain why a behavior that is reinforced only occasionally can persist: it continues as long as its relative payoff compared with the alternatives is high enough.

In practice: In class, a student's off-task joking produces peer laughter about every two minutes, while working produces teacher praise about every ten minutes; she jokes much more than she works. Increasing praise for work and arranging for peers to ignore the jokes shifts her allocation toward work, as the matching law predicts.

Where questions turn: The matching law describes relative allocation among alternatives available at the same time; it predicts proportions of responding, not the absolute rate of a single behavior under one schedule.

Practise B.23

B.24Identify and distinguish between imitation and observational learning.

Imitation is behavior that meets four conditions: a model's behavior is the antecedent; the imitative behavior follows within a short time; it has formal similarity to the model (looks or sounds like it); and the model is the controlling variable for it.
Formal similarity alone is not imitation: two people who both open umbrellas because it starts raining are responding to the rain, not to each other.
Generalized imitation is imitating new models, never specifically reinforced, after a history of reinforcement for imitating many other models; it is a prerequisite for learning by modeling.
Observational learning is broader: learning that results from watching another person's behavior and its consequences; the observer's behavior may occur much later and in another setting, and it often depends on seeing the model reinforced or punished.
Observational learning usually requires component skills: attending to the model, imitating, and discriminating the consequences the model received.

In practice: A child copies his therapist's clapping right after she claps (imitation). The next week, having watched a classmate earn a sticker for putting toys away, he begins putting toys away himself at cleanup time without being shown or told, which is observational learning.

Where questions turn: Imitation requires that the model controls an immediate, formally similar response; delayed behavior acquired by watching someone else's consequences is observational learning.

Practise B.24

Key terms

Response class: A group of responses that share a common function or form.
Positive reinforcement: A stimulus is added after a response and the response increases in future.
Negative reinforcement: A stimulus is removed after a response and the response increases in future.
Positive punishment: A stimulus is added after a response and the response decreases in future.
Negative punishment: A stimulus is removed after a response and the response decreases in future.
Generalized conditioned reinforcer: A conditioned reinforcer paired with many other reinforcers, effective across motivational states.
Postreinforcement pause: The absence of responding after reinforcement, typical of fixed ratio and fixed interval schedules.
Extinction burst: An initial increase in the frequency, duration or intensity of behavior when reinforcement is first withheld.
Spontaneous recovery: The reappearance of an extinguished behavior after a period without it, while extinction remains in place.
Discriminative stimulus (SD): A stimulus in whose presence a response has been reinforced, which evokes that response.
Motivating operation: An environmental variable that alters the value of a consequence and the current frequency of behavior that produced it.
CMO-T: A conditioned motivating operation that makes another stimulus effective as a reinforcer.
CMO-R: A conditioned motivating operation that signals worsening or improvement and makes its own removal or presence a reinforcer.
Mand: A verbal operant evoked by an MO and reinforced by the specific consequence it specifies.
Tact: A verbal operant evoked by a nonverbal stimulus and maintained by generalized conditioned reinforcement.
Intraverbal: A verbal operant evoked by a verbal stimulus without point-to-point correspondence.
Stimulus equivalence: Emergence of untrained reflexive, symmetrical and transitive relations among stimuli after training some relations.
Resistance to change: The persistence of behavior when a disruptor such as extinction is applied.
Matching law: On concurrent schedules, the relative rate of responding matches the relative rate of reinforcement.

After this guide: Given a brief episode, you should be able to classify the contingency, schedule, reinforcer or punisher type, antecedent function, verbal operant or learning process it shows, and say which feature of the episode decides the classification.

Practise area B