When to use it
Use this sheet to teach and track any behavior chain, a skill made up of a sequence of steps such as handwashing, dressing, packing a backpack, or making a snack. Each step is listed separately, and for each session you record the level of prompt the learner needed on that step.
The sheet works with the three common chaining methods:
- Forward chaining: teach the first step until it meets criterion, then add the second, and so on. The instructor completes (or fully prompts) the remaining steps.
- Backward chaining: the instructor completes (or fully prompts) every step except the last. The learner does the last step and immediately contacts the natural end of the task. Once the last step is mastered, teach the second-to-last, and so on. This is often used when finishing the task is itself reinforcing.
- Total-task chaining: teach every step in every session, prompting as needed. This is often a good choice when the learner already has some of the steps or when the chain is not long.
No single method works best for every learner or skill. The choice depends on the length of the chain, the learner’s current skills, and what motivates them. Write the method on the sheet.
How to fill it in
- Write the task analysis first. Break the skill into small, observable steps. Build it by watching a competent person do the task, doing it yourself, or asking the family how they do it at home. Steps should match the learner’s routine and materials.
- Assess before teaching. Run a baseline probe. In a single-opportunity probe, stop at the first step the learner can’t do independently and score the rest as not independent. In a multiple-opportunity probe, complete any missed step for the learner (or set it up) and let them try each remaining step. Note which method you used, because the scores aren’t directly comparable.
- Write the chaining method and prompt strategy at the top of the sheet. Common prompt strategies include most-to-least, least-to-most, graduated guidance, and prompt delay.
- For each session, write the date at the top of a column. Then record the prompt level for every step taught, using the key below. Code the least intrusive prompt that was needed for the learner to complete the step.
- Mark steps that were not taught this session (for example, the steps the instructor completes in forward or backward chaining) with NT. This keeps them from being scored as errors.
- Total the column: count steps scored I and divide by the number of steps taught or probed.
Prompt-level key (most to least intrusive; adjust to match the learner’s program):
| Code | Meaning |
|---|---|
| FP | Full physical: hand-over-hand guidance through the whole step |
| PP | Partial physical: light touch or brief guidance, such as at the wrist or elbow |
| M | Model: instructor demonstrates the step |
| G | Gestural: pointing, nodding, or motioning toward the next step |
| V | Verbal: a spoken cue or instruction for the step |
| I | Independent: completed correctly with no prompt |
| - | Incorrect or no response, even with the prompts available |
| NT | Not taught or not probed this session |
Where verbal and gestural prompts rank relative to each other varies across programs and learners. Verbal prompts can be harder to fade for some learners. Use the order written in the learner’s program.
Common mistakes:
- Steps that are too big, such as “brush teeth” as one step. If a learner gets stuck on a step, break it into smaller steps.
- Scoring a step I when the learner needed a small gesture or reminder.
- Adding verbal cues to physical prompts without writing them down, which creates hidden prompt dependence.
- Changing the order or number of steps without starting a new sheet or marking a phase change.
The sheet
Client:Skill:Instructor:Setting:
Chaining method (forward / backward / total task):Prompt strategy:
Mastery criterion:Probe type (single / multiple opportunity):
| Step # | Step | Date: | Date: | Date: | Date: | Date: | Date: | Date: | Date: |
|---|---|---|---|---|---|---|---|---|---|
| 1 | |||||||||
| 2 | |||||||||
| 3 | |||||||||
| 4 | |||||||||
| 5 | |||||||||
| 6 | |||||||||
| 7 | |||||||||
| 8 | |||||||||
| 9 | |||||||||
| 10 | |||||||||
| 11 | |||||||||
| 12 | |||||||||
| 13 | |||||||||
| 14 | |||||||||
| 15 | |||||||||
| Steps independent | |||||||||
| Steps taught / probed | |||||||||
| % independent | |||||||||
| Initials |
Summarizing the data
Percent of steps independent = steps scored I divided by steps taught or probed that session, times 100.
Worked example: a 10-step total-task chain. In one session, the learner completed steps 1, 2, 3, 5, 9, and 10 independently. They needed a gesture on step 4, a model on steps 6 and 7, and partial physical guidance on step 8.
- Steps independent: 6
- Steps taught: 10
- Percent independent = 6 / 10 x 100 = 60%
In forward or backward chaining, also track progress on the current training step by itself. For example, the mastery criterion might be “the training step is independent in 3 consecutive sessions before adding the next step.” The percent-independent figure for the whole chain will rise in steps as each new step is added.
Some teams also give each prompt level a score (for example, I = 5 down to FP = 0) so they can graph gradual prompt fading before steps become fully independent. If you do this, keep the scale identical across sessions and staff.
Tips
- Review the steps that are hardest for the learner, not just the total. A total that stays flat at 80% often means one or two stuck steps need to be broken down, taught differently, or prompted another way.
- Teach the chain at the times and places it naturally happens, such as handwashing before snack, whenever possible. This helps the skill generalize.
- Use the same materials the learner will use at home, or plan to generalize to them.
- Once the chain is mastered, probe it on a maintenance schedule, and check whether the learner can do it in a new setting or with a new person.
- For skills with safety risks (hot water, knives, stoves, crossing streets), write safety rules into the task analysis and have the supervising BCBA approve the level of supervision.