Translated from the Portuguese original. Read the original (Português)
Chaining is the procedure used to teach a sequence of behaviors — brushing teeth, getting dressed, making a snack, catching the bus. First the task is broken into small, observable steps (the task analysis); then these steps are linked to one another until the routine works from start to finish. The secret lies in the joints: each completed step changes the environment, and that change reinforces the step that just happened while, at the same time, signaling what comes next.
A behavior chain is a specific sequence of discrete responses, each associated with a particular stimulus condition. Completing each link produces a stimulus change with a dual function: it serves as a conditioned reinforcer for the response that just occurred and as a discriminative stimulus (SD) for the next response — and it is this dual function that holds the chain together. Only the last response produces the terminal reinforcer. Chaining is the set of procedures that links these links together, starting from a task analysis and, usually, with the support of prompting and fading (Cooper, Heron & Heward).
Why teach in links
Much of what matters in everyday life is not one behavior: it is a series of them, in an order that matters. Taking a shower, brushing teeth, packing a backpack, crossing the street, making a sandwich — in each of these cases, knowing the part is not enough; the parts must be linked in the right sequence, from beginning to end.
Asking "brush your teeth" of someone who has never brushed is asking for a whole chain at once. The result tends to be predictable: the person does not finish, does not contact any good outcome, and the task becomes a source of frustration. Chaining turns this problem inside out — instead of demanding the whole, it teaches link by link, with reinforcement guaranteed along the way.
- the steps, on their own, are simple, and the person often already does several of them;
- teaching focuses where the real difficulty is, not on the whole routine;
- the sequence becomes measurable: it is possible to record exactly which step the person stopped at;
- once the chain is solid, the task itself starts to "pull itself along" — one step summons the next, with no one prompting.
The anatomy of a chain
The most interesting part of a chain is not the steps: it is the joints between them. When someone picks up the toothbrush, the world changes a little — now there is a brush in hand. That change does two things at once: it gives an "okay, that worked" for the step that just happened and signals what the next step is (putting on toothpaste). That is why the chain does not fall apart in the middle.
Chaining is not shapingShaping builds one behavior that does not yet exist, by reinforcing successive approximations. Chaining links in sequence distinct behaviors — generally already present in the repertoire, or taught before, one by one. Shaping gives form; chaining connects. In practice, the two often appear together: the step the person cannot yet do is shaped, and the whole routine is chained.
Before teaching: the task analysis
No chain is taught without first writing the task analysis: the list of steps, in order, in observable language ("turn on the faucet," not "take care of hygiene"). Steps that are too large hide the difficulty; steps that are too small make teaching endless. There are three classic ways to build and validate this list:
- observe someone who already does the task well and write down each step in the order it happens;
- consult specialists or already published and validated task analyses for that skill;
- perform the task yourself, recording what you do — this is how the "invisible" steps that no one remembers to list show up.
With the list ready, it remains to find out where the person starts from: which steps they already perform alone. This chain assessment has two known forms. Tap the cards:
The three chaining methods
With the steps and the starting point defined, what remains is choosing where to start linking. There are three main procedures — and one useful variation. Explore each one:
Forward chaining
Teaching happens in the natural order, from the first step forward. Step 1 is trained to criterion, with reinforcement upon completing it; the remaining steps are handled by the teacher or are not yet trained. Once step 1 is mastered, reinforcement comes to depend on 1 + 2, then 1 + 2 + 3, and so on.
In its favor: the person learns the sequence in logical order, from beginning to end, and each phase adds little that is new.
Backward chaining
The teacher performs all steps except the last; the learner performs that last step and immediately contacts the terminal reinforcer — the finished task. Once the last step is solid, the learner moves on to performing the last two, then the last three, until doing the whole chain.
In its favor: from the very first trial the person completes the task and contacts its natural outcome — the shirt put on, the snack ready — which makes the effort understandable and worthwhile.
Total task presentation
On every trial the learner goes through all the steps of the chain, receiving prompts (for example, graduated guidance) only on the steps not yet mastered, until the whole chain becomes independent. There is no need to master one step before moving to the next.
In its favor: the routine is practiced whole, without cutting it up, and the steps already mastered continue to be performed independently — which brings teaching closer to real life.
Backward chaining with leap aheads
A variation on backward chaining for someone who already knows part of the task: the steps that the assessment showed were already in the repertoire are not formally trained — they are only probed from time to time. Training focuses on the links that are missing.
In its favor: it shortens the path considerably and avoids teaching what has already been learned, favoring faster independence.
Which method is best?None of the three is superior in every situation: reviews comparing forward, backward, and total task chaining have found mixed results. The choice is clinical and made case by case — it depends on what the learner already masters, the length and complexity of the task, how much support is available, and the context in which the routine will happen. Deciding based on the learner's own data, rather than a preference for the method, is what sustains evidence-based practice.
When the chain falls apart
Chains that used to work can get stuck, almost always for identifiable reasons. It is worth reviewing:
- is the task analysis complete? A step forgotten from the list is often exactly the point where the sequence stops;
- is the chain too long or too complex for the moment? Long sequences call for more support and, sometimes, splitting into blocks;
- how is the reinforcement schedule? Requiring the whole chain for too long with nothing along the way weakens the early links;
- was there stimulus variation? A different bathroom, a different sink, a different brush — the SD changed, and stimulus control may not have generalized;
- was there response variation? A faucet that turns instead of lifts requires a different response from the one trained.
There is also a quiet risk: prompt dependence. If help is never withdrawn, the person learns to wait for the cue at each step instead of being guided by the effect of the previous step itself — and the chain only "moves" with someone alongside. That is why a fading plan is part of teaching from day one. Teaching a self-care routine is also a decision about someone's life: choosing tasks the person themselves values, asking for their assent, and respecting their pace is part of the procedure, not a side detail.
Test your understanding
Quick quiz
Three questions to lock in the concept. You see the explanation right away.
Further reading
- Cooper, J. O.; Heron, T. E.; Heward, W. L. Applied Behavior Analysis. 3rd ed., 2020 — chapter on Chaining.
- Spooner, F.; Spooner, D. "A Review of Chaining Techniques: Implications for Future Research and Practice". Education and Training of the Mentally Retarded, v. 19, n. 2, p. 114–124, 1984.
- Slocum, S. K.; Tiger, J. H. "An assessment of the efficiency of and child preference for forward and backward chaining". Journal of Applied Behavior Analysis, v. 44, n. 4, p. 793–805, 2011.
- Behavior Analyst Certification Board (BACB). Test Content Outline (6th ed.), item G.12 — select and implement chaining procedures.
This entry is educational and does not replace assessment, diagnosis or treatment by a qualified health professional. Every person is unique — for a specific case, seek individual guidance.