How Should ABA Chaining Data Be Reported?

Structure ABA chaining data by step

Use one row per eligible step or another format that preserves the same information. Define independent before collection. Keep prompts by level, requested help, partner action, no response, errors, withdrawal, and invalid materials separate. Mark the active forward, backward, total-task, or modified method.

Use matching denominators

Step independence equals independent steps divided by eligible learner steps. Whole-routine completion equals completed routines divided by started eligible routines. Fidelity equals correctly implemented teaching actions divided by actions due. Report duration with start and end events.

Keep versions apart

A reordered task, new tool, different prompt, changed completion arrangement, or new setting may require a new series. The Ethics Code supports documentation and data-based evaluation for covered people. Preserve prior data and pool only when definitions, methods, and conditions support a valid comparison.

Interpret studies cautiously

Bancroft and colleagues found participant- and measure-specific efficiency among forward-chain completion procedures. The BACB outline covers measurement, chaining, integrity, generalization, maintenance, and evaluation. Neither supplies a universal mastery percentage.

A practical example

Across two seven-step café-closing routines, Sora completes eight steps independently, two after gestures, requests help once, has one partner-completed safety step, and encounters two invalid supply steps. Report each count, both routine outcomes, and the method version rather than 8 of 14 independence.

Questions families can use

Which steps were eligible for the learner? Who completed the rest? Are help and prompts separate? Which events were invalid? Did definitions or methods change? What did the person report?