What Happens When an ABA Reinforcer Stops Working?

Review the changed effect

Compare recent and earlier periods using the same response and opportunity definition. Record consequence due and delivered, client acceptance, latency, independent response, prompts, setting, partner, health changes, competing activities, and preference check. Choose one repair and a review period.

Begin by replacing “stopped working” with observable facts. State the response, current opportunity count, earlier and recent response distribution, prompt levels, consequence, delivery integrity, and client acceptance. A decrease from eight of ten to two of ten means something different if the task, staff, or eligibility rule also changed.

Review the sequence in order:

1. Preference: Does the person currently choose or accept the event among meaningful alternatives?
2. Delivery: Did the planned event follow the eligible response, in the right amount and within the defined time?
3. Task and cue: Did response effort, materials, natural cue, prompt timing, or setting change?
4. Competing outcomes: Is another activity or natural consequence now more valuable?
5. Health and access: Did pain, fatigue, medication, illness, sensory conditions, AAC, motor access, or another need change?
6. Goal: Does the person still want and benefit from the target response in this context?

Each question has a different repair. A delivery problem calls for implementation correction. A preference change calls for current choice. An inaccessible task may need adaptation or another professional. A goal that no longer matters may need to end.

Keep the data streams separate

Use separate measures for preference selections, consequences due and delivered, client acceptance, target responses, prompt levels, and adverse effects. Do not call a low target-response rate proof that the event is no longer preferred. Do not call a high selection rate proof that it still strengthens the target response.

Preserve invalid opportunities and missed delivery. Removing them can make the clinical outcome look better while hiding a system problem. Segment by plan version, staff, and setting when meaningful, with raw counts alongside percentages.

Define what the evidence means

The contingency may be unclear or delayed. The response may require more effort than the outcome is worth. Another natural consequence may now matter more. The goal itself may also have lost value. Each possibility leads to a different clinical question.

Avoid making the event scarce, increasing deprivation, or withholding basic needs to restore its effect. These actions can create coercion and health risk. Communication, food, water, bathroom access, mobility, prescribed care, pain care, rest, safety, and ordinary relationships require appropriate access independent of performance.

The client should be able to choose another option or decline. When distress, sudden aversion, pain, swallowing difficulty, allergic signs, or another health concern appears, stop the arrangement and use the appropriate health or emergency route. Outcome measurement should not delay care.

Test one relevant repair

Predeclare the repair, ordinary supports, opportunity count, measures, stop conditions, and review date. Examples include restoring timely delivery, offering a current choice, reducing task effort, clarifying the natural cue, adapting materials, or shifting to a natural outcome. If several changes are needed for comfort or access, make them and report a bundled change without attributing the result to one element.

Pair performance with client feedback and unwanted effects. More target responses do not by themselves establish that the revised arrangement is acceptable or meaningful.

Use current clinical and ethics sources

The CASP public summary places assessment, planning, implementation, and evaluation within its autism-treatment scope. The BACB Ethics Code addresses client involvement, consent and assent when applicable, assessment-based intervention, positive reinforcement, risk, documentation, and evaluation for covered behavior analysts.

The BCBA Test Content Outline covers preference assessment and positive and negative reinforcement procedures as examination content. It does not make a preferred event a reinforcer or prescribe one outcome, schedule, or assessment for every person.

Use preference assessments as candidate evidence

Lill, Shriver, and Allen synthesized 65 articles into SPADS to help trained practitioners choose context-specific stimulus-preference assessments. The assessment identifies candidates. Later response data determine whether an event functioned as reinforcement in the defined context.

Keep communication and refusal available

The ASHA AAC portal supports continuous communication-tool access. The client needs a reliable way to choose, decline, pause, change an option, report discomfort, and request basic needs throughout assessment and teaching.

A practical example

A puzzle break previously followed eight of ten independent work starts. In a recent period, independent starts are two of ten. The plan and task appear unchanged, but the client declines the puzzle after both starts and chooses conversation instead of puzzles in five of six preference presentations.

The team can say current preference appears different among the tested options. It cannot yet conclude that preference change caused fewer work starts. The clinician also checks prompt timing, task effort, delivery records, health, and whether the work goal still matters.

With the client's agreement, the next ten comparable opportunities offer a choice of conversation, puzzle, or decline after a start. Independent starts occur in five; conversation is selected after four, and decline after one. These small sequential samples support continued review. They do not prove that conversation reinforced starting or predict another task.

Review changed-preference and stopping questions

Ask which fact triggered the review, whether preference and delivery were checked separately, what changed in the task or setting, and what the client reports. Request one proposed repair with a defined recheck rather than a broad increase in rewards or practice.

Also ask what would end the arrangement. A plan should stop or change when the client withdraws, the consequence becomes unavailable or unsafe, health concerns arise, burden outweighs benefit, or the goal no longer fits. Document the clinical decision and effective version before routine implementation.

Bring a short timeline with the last period in which the arrangement appeared useful, the first observed change, and any changes in health, staff, setting, materials, schedule, ordinary access, or client preference. Include specific counts rather than a general statement that motivation dropped.

Assign owners to different problems. A supervisor may address missed delivery, operations may repair equipment, an appropriate health professional may assess pain or swallowing concerns, and the qualified clinician may revise the clinical plan. Keep each action and due date visible.

After the repair, preserve stopped, declined, and invalid opportunities. Report whether implementation improved, whether the person still chose the event, and whether later responding changed. These related questions retain separate answers.

Questions families can use

Ask whether preference changed, how recently the event was available, whether delivery matched the plan, what effort the response requires, what health or context changed, and whether the goal still matters.