Task analysis

The sequence belongs to the person and setting

Two people can complete the same activity through different movements, tools, orders, and supports. Build the sequence around the person's goals and effective method. Account for mobility, vision, hearing, sensory access, communication, health, handedness, culture, and available materials.

Avoid treating one staff member's habit as the correct sequence. Some steps can occur in more than one safe order. Mark flexible and required order explicitly.

Write observable step boundaries

Each step needs a clear start and completion. “Gets ready” hides several actions. “Places gloves on both hands using the chosen technique” gives an observer more information.

Use language that permits acceptable variation. Define materials, relevant position, safety condition, and ordinary support. Keep the list detailed enough to teach and measure while avoiding tiny divisions that make the routine burdensome.

Validate the analysis in real use

Observe the task in the actual setting with ordinary materials. Ask the person to review or demonstrate when appropriate. Consult relevant family members and professionals for health, motor, communication, sensory, or safety needs.

Walk through every step before teaching. Check missing prerequisites, repeated steps, inaccessible materials, transitions, cleanup, and natural outcomes. Revise the version number and date when the sequence changes.

The task analysis is separate from teaching

The step list describes the routine. Forward chaining, backward chaining, total-task presentation, modeling, prompting, reinforcement, and error correction describe teaching choices. The written sequence alone does not select among them.

The ABAI basic-principles page and BACB Test Content Outline cover task analysis, chaining, prompting, fading, generalization, maintenance, and evaluation as educational or examination content.

Choose the measurement unit

Step-level measurement shows which parts are independent, prompted, incorrect, skipped, or invalid. Whole-chain measurement shows whether the complete routine reaches its functional outcome. Both can be useful.

Define whether a prompted step counts as completed and report independence separately. Record environmental failures such as missing materials. An unavailable tool should not become a learner error when providing it belongs to the partner or setting.

Version the sequence and its data

Give every task analysis an owner, effective date, version, setting, materials list, and review trigger. When a step changes, preserve the prior version and mark which sessions used each one. Data collected under different sequences should remain distinguishable.

Test the revision with the person before wider use. A shorter list may combine actions that need separate teaching, while a longer list may add unnecessary burden. Compare the new version for access, safety, time, independence, and preference. Update staff examples, data forms, visual supports, and training together. If one site or caregiver uses another safe method, document the variation rather than scoring it as incorrect. The record should show whether performance changed or the definition changed.

When materials differ across settings, link each variation to the same version or create a scoped local version. This prevents a harmless adaptation from looking like poor fidelity. Review both.

A fictional garden example

Dev chooses a seven-step watering routine at a community garden. Across six complete routines, 42 step opportunities occur. Dev completes 34 of 42 steps independently and finishes the whole routine without added prompts in 3 of 6 opportunities.

The two denominators answer different questions. One routine stops because the water line fails; under the prewritten rule, its observed steps remain described while the whole-chain opportunity is marked invalid. These counts do not prove that the teaching procedure caused progress.

Protect assent, access, and safety

Offer an accessible way to accept, decline, pause, or change the routine. Explain any physical contact before it occurs and honor withdrawal. A task analysis does not authorize restraint, forced practice, or removal of essential support.

ASHA's AAC guidance says AAC users should always have their tools or devices. Communication, mobility, prescribed care, food, water, bathroom access, rest, pain support, and emergency help remain available.

Revise from performance and experience

Track step and chain performance, prompt levels, time, errors, withdrawal, partner fidelity, generalization, maintenance, burden, and the person's view. A long pause at one step may indicate unclear materials or an inefficient sequence.

The current BACB Ethics Code addresses client involvement, consent and assent when applicable, assessment, positive reinforcement, risk, and ongoing evaluation for covered professionals. Update the analysis when the task, setting, ability, support, or preference changes.