What is the difference between DRO and DRA?

DRA reinforces a named alternative behavior, ideally one with the same function as the problem behavior. DRO reinforces the absence of the problem behavior for an interval and names no replacement. A plan that says what to do instead is DRA; one that only sets a timer is DRO.

What is the difference between shaping and chaining?

Shaping changes one response gradually by reinforcing successive approximations (G.8). Chaining links several different responses from a task analysis into a sequence (G.9), taught forward, backward or as a total task.

When is most-to-least prompting used?

For new skills, and for learners whose errors tend to repeat: start with the most intrusive prompt that secures a correct response and fade toward the natural cue.

Is noncontingent reinforcement the same as DRO?

No. Noncontingent reinforcement arrives on a time schedule whatever the learner does; DRO arrives only if the target has been absent for the interval. NCR sits under task G.1, DRO under G.18.

What is the heaviest section on the BCaBA exam?

Section G, Behavior-Change Procedures: 22 of the 150 scored items, about 15%, under the 6th-edition outline per the current BCaBA Handbook (updated June 2026). Sections C and H follow with 20 each.

Are schedules of reinforcement tested in Section G?

Their definitions sit in Section B (B.8), and the BCaBA outline covers simple schedules only. Section G uses them inside procedures: thinning to an intermittent schedule for maintenance, time-based delivery in NCR, and rate limits in DRL.

How many Section G practice questions are on this site?

Twelve on this page and nine more under the G chip of the 60-trial practice exam; the two sets do not overlap.

Section guide

Section G for the BCaBA: behavior-change procedures, the outline's heaviest section

Section G, Behavior-Change Procedures, is the heaviest section on the BCaBA® (Board Certified Assistant Behavior Analyst) exam: 22 of the 150 scored items, about 15%, under the Test Content Outline (6th ed.), the document formerly called the task list and in effect since 2025. Its twenty tasks, G.1 to G.20, cover how you build behavior, how you teach it, how you reduce it and how you make it last. This page walks them in that order, names the look-alike answers in each group, and ends with 12 practice questions.

Get the appBCaBA practice on iPhone Practice free on this site

What Section G covers on the BCaBA (22 scored items)

Under the 6th-edition outline, per the current BCaBA Handbook (updated June 2026), the exam has 150 scored items across nine sections, and Section G supplies 22 of them, more than any other. Sections C and H come next with 20 each. The full nine-section map, with every weight, is on the outline hub.

22scored itemsof 150, 6th-ed. outline

15%of the scored test

20tasksG.1 to G.20

12practice trialsfurther down this page

Twenty-two items over twenty tasks is just over one item a task, so no single procedure carries the section: breadth beats depth. The table regroups the tasks by the decision a question asks you to make.

Section G's twenty tasks, regrouped by the decision in the stem

Group	Tasks	The decision you make
Building behavior	G.1 to G.11	Which reinforcer, which cue, which prompt, and how the help fades
Teaching arrangements	G.13 to G.16	Who starts the trial, and who holds the contingency
Reducing behavior	G.18, G.19	What earns reinforcement, and what is added or removed after the behavior
Making it last	G.12, G.17, G.20	Does it hold in another setting, in another form, at a later date

Few Section G items ask for a bare definition. The usual stem describes a session (a learner, a cue, what the implementer does next) and asks you to name the procedure or pick the next step. Name it in your own words before you read the four options.

Day 1: the 2×2 grid
Positive or negative, reinforcement or punishment. The grid is Section B material, but most wrong options in Section G are a wrong cell of it.
Days 2 and 3: reduction
The differential-reinforcement family and punishment (G.18, G.19), where the look-alike options cluster.
Days 4 and 5: building
Reinforcers, prompts and fading, shaping, chaining, verbal operants and emergent relations (G.1 to G.11).
Day 6: arrangements and durability
Teaching arrangements, group contingencies, contracts, generalization and maintenance.
Day 7: trials
The 12 trials below untimed, misses reworked the next day.

Building new behavior: reinforcement, prompts, shaping and chaining

Building items describe a learner who cannot yet do something. Sort each stem into one of four questions: what is the reinforcer, what signals it, what helps the response happen, and how does that help leave.

Reinforcers and the conditions around them (G.1 to G.4)

Contingent reinforcement (G.1): Delivered after a response. Positive adds a stimulus, negative removes one; both increase the behavior.
Noncontingent reinforcement (G.1): The maintaining reinforcer arrives on a fixed-time or variable-time schedule whatever the learner does, which lowers the motivation behind the problem behavior.
Conditioned reinforcer (G.2): Gains value by pairing with reinforcers that already work. A token exchanged for several backup reinforcers is a generalized conditioned reinforcer.
Motivating operation (G.3): Changes a reinforcer's value right now: deprivation raises it, satiation lowers it.
Discriminative stimulus (G.3): Signals that reinforcement is available; it leaves the value alone.
Simple and conditional discrimination (G.4): Simple: one cue, one correct response. Conditional: the correct choice depends on a sample, as in matching-to-sample.

Prompts and how they fade (G.5 to G.7)

Response prompts act on what the learner does: verbal, gestural, model, physical. Stimulus prompts change the cue itself, making the correct card bigger, brighter or closer. Modeling (G.6) and instructions (G.7) have their own task codes but behave like response prompts: added help that has to leave.

A response-prompt ladder, most to least intrusive

Rung	Prompt	What the implementer does
1	Full physical	Guides the learner's hands through the whole response
2	Partial physical	A light touch at the wrist or elbow to start it
3	Model	Performs the response for the learner to imitate
4	Gestural	Points, nods or looks toward the correct choice
5	Verbal	Says what to do, or the first sound of the answer
6	Independent	Nothing: the natural cue alone

Texts differ on where gestural and verbal sit; the two ends never move.

Most-to-least

Starts at the most intrusive rung that secures a correct response.
Moves down the ladder as the learner succeeds.
Few errors; suits new skills and learners whose errors tend to repeat.

Least-to-most

Starts with the natural cue and a chance to respond alone.
Climbs a rung only after no response or an error.
More errors; suits responses the learner can already partly do.

The tellLook at the first thing the implementer does on a trial. Full help up front means most-to-least; waiting for an independent attempt means least-to-most.

Time delay is the third fading method. The prompt first comes 0 seconds after the cue, then later: in one jump (constant, for example 0 to 4 seconds) or a second at a time (progressive). If the stem counts seconds between cue and prompt, it is time delay, not a ladder.

Shaping or chaining (G.8, G.9)

Shaping

One response, changed gradually.
Reinforce each closer approximation and stop reinforcing the earlier ones.
Can target a dimension: duration, force, latency.

Chaining

Several different responses linked in order.
Built from a task analysis.
Finishing one step is the cue for the next.

The tellCount the behaviors. One behavior moving toward a criterion is shaping; a list of different steps is chaining.

Three ways to train a chain

Method	Taught first	Where reinforcement lands
Forward	Step 1, then steps 1 and 2, and so on	After the step being taught
Backward	The last step; the implementer completes the earlier ones	At once: the learner finishes the chain on every trial
Total-task	Every step on every trial, prompted where needed	After the whole chain

A worked case: a six-step task analysis for making instant oatmeal. Under backward chaining the implementer completes steps 1 to 5 and the learner stirs and takes the first spoonful, so the terminal reinforcer comes at once. When that step is independent, the learner takes over step 5 too.

Verbal operants and emergent relations (G.10, G.11)

Mand: a request under a motivating operation, reinforced by the specific thing asked for. Teach it when the learner wants the item.
Tact: naming something in view, reinforced by praise or attention. No MO needed.
Echoic: repeating a spoken model, with point-to-point correspondence and the same form.
Intraverbal: answering a verbal cue with different words, such as answering a question.
Emergent relations: train A to B and B to C, then test B to A (symmetry) and A to C (transitivity) without training them. Emergent means untrained.

Choosing how to teach: discrete-trial, free-operant, naturalistic

Task G.13 asks you to match the teaching arrangement to the client and the skill. Start with one question: who starts each learning opportunity?

Three acquisition arrangements

Arrangement	Who starts it	Fits when
Discrete-trial	The implementer: cue, response, consequence, intertrial interval	A new skill needs many fast, countable learning opportunities
Free-operant	The learner, who can respond at any time, so rate means something	The skill has to become fluent at a natural pace
Naturalistic (incidental)	The learner, inside a routine, using motivation already present	The skill must carry into everyday settings, such as requests at snack time

Group contingencies (G.14)

Group contingencies: whose behavior decides the reward

Type	The criterion applies to	Who earns the reward
Independent	Each person separately	Each person who meets it
Dependent	One person or a small set	The whole group
Interdependent	Every member, or the group total	The whole group, only if the criterion is met

The middle column settles it. A class that earns extra recess when one student meets a reading goal is dependent (the hero procedure); a class that earns it only if every student does is interdependent.

Contracts and self-management (G.15, G.16)

A contingency contract names the task, the reward, who does what, by when and to what criterion, signed by both parties. Self-management puts the person in the implementer's seat: self-monitoring, self-instruction, self-delivered consequences. Self-recording alone often changes behavior through reactivity, a measurement problem you meet again in Section C, Measurement, Data Display, and Interpretation. The classic failure is short-circuiting: taking the reward without meeting the criterion.

Reducing behavior: the differential-reinforcement family and punishment

Reduction items hold Section G's closest look-alikes: five procedures that all lower a problem behavior and differ only in what earns reinforcement. Before the options, finish one sentence: reinforcement is delivered when ___. The path below fills the blank.

Does the reinforcer arrive on the clock, whatever the learner does?
Yes: noncontingent reinforcement, fixed-time or variable-time (G.1). No: go on.
Does the plan name a behavior to reinforce?
Yes: DRA, next step. No: skip to the absence question.
Can that behavior and the problem behavior happen at the same moment?
No: DRI. Yes: plain DRA, strongest when functionally equivalent, such as a request for a break replacing escape-maintained aggression.
Is reinforcement delivered when the target has been absent for an interval?
Yes: DRO.
Is some of the target fine, as long as it stays under a limit?
Yes: DRL (full-session, interval or spaced-responding). The goal is a lower rate, not zero.

The family side by side (G.1, G.18)

Procedure	What earns reinforcement	Goal for the target
DRA	A named alternative behavior	Replaced
DRI	An alternative that cannot occur at the same time	Replaced
DRO	Absence of the target for an interval	Zero within each interval
DRL	The target at or under a set rate, or spaced out	A lower rate
NCR	Nothing the learner does; time passing	Less motivation to emit it

Noncontingent reinforcement

Time-based: fixed-time or variable-time.
If the target occurs, the reinforcer still comes on time.
Works mainly as an abolishing operation.

DRO

Delivered only if the target was absent for the interval.
A target response usually resets the interval.
Works through differential reinforcement.

The tellAsk what happens if the target occurs one second before delivery. Under NCR the reinforcer still arrives; under DRO it does not.

Setting a first DRO interval from baseline (worked example)

Baseline session30 minutes
Target responses counted10
Average time between responses30 / 10 = 3 minutes
Start a little under thatfor example 2 minutes

First DRO interval: 2 minutes

Hypothetical numbers. Start with an interval the learner can meet; lengthen it as the target stays low.

With and without extinction

Task G.18 reads "with and without extinction". With extinction, the problem behavior stops producing its maintaining reinforcer; without it, both behaviors still pay and the alternative has to out-compete the problem. Extinction is defined by function: withholding attention works only for attention-maintained behavior, and escape extinction keeps the demand in place.

DRA plus extinction for attention-maintained shouting (hypothetical)

Hypothetical data. The first sessions of the second phase rise above baseline: an extinction burst. Dropping extinction during the burst teaches that louder shouting pays. Side effects as such are task H.5.

Punishment procedures (G.19)

Three punishment procedures in the outline

Procedure	Type	What follows the behavior
Time-out from positive reinforcement	Negative	Access to reinforcement removed for a period, exclusionary or non-exclusionary (planned ignoring, contingent observation)
Response cost	Negative	A set amount of an earned reinforcer, such as tokens, is taken back
Overcorrection	Positive	Effortful behavior added: restitution or positive practice

Making change last: generalization, maintenance and persistence

Three tasks ask whether a change holds: generalization (G.17), maintenance (G.20) and persistence (G.12). Another setting, another form, a later date, a disrupted routine.

Stimulus generalization

The trained response appears under untrained conditions.
New people, rooms, materials, wording.

Response generalization

Untrained responses appear that do the same job.
A learner taught to say hello also starts waving.

The tellAsk what is new: the conditions (stimulus) or the behavior (response). If nothing is new except the date, it is maintenance.

The tactics stems point to come from Stokes and Baer (1977). Each one is a concrete change to training:

Teach with several teachers, rooms and materials (sufficient exemplars).
Train loosely: vary the wording, timing and position of cues.
Bring a stimulus from the target setting into training (common stimuli).
Make the contingencies hard to tell apart: intermittent or delayed reinforcement.
Hand the behavior over to reinforcement the setting already provides (natural contingencies).

Maintenance items usually point at schedule thinning: continuous to intermittent reinforcement before the program ends, because intermittently reinforced behavior resists extinction better.

Behavioral momentum is the usual G.12 procedure: three to five quick, easy requests the learner reliably completes, each reinforced, then the harder one. Behavior reinforced at a high rate in a setting is harder to disrupt there.

Five Section G traps

These five turn up as rules in study notes; each one costs points in more than one task.

One habit covers all five: translate the stem into added or removed, and up or down, before you read a single option.

12 Section G practice questions

Twelve Section G trials follow, one at a time, each written for this site to a G task of the 6th-edition outline. The trial sheet above the card marks + for a correct answer and a dashed minus for an error; the cumulative record beside it climbs one step per correct answer and runs flat on an error. Every option gets a note. Nothing is stored, and the record predicts no exam result.

12 Section G trials

Answer with 1–4 or A–D; the arrow keys step between trials

Trial sheet · 12 trials

Cumulative record: 0 correct over 0 answered

Trial 1 of 12

GBehavior-Change ProceduresG.4

When teaching a conditional discrimination where a learner selects the matching picture from an array, what must occur for the discrimination to be established?