Errorless teaching data should separate independent correct responses, prompted correct responses by level, errors, no response, self-correction, client withdrawal, and invalid opportunities. Add prompt timing, staff fidelity, latency, duration, and later unprompted probes. High prompted accuracy shows successful supported responding under that procedure. It does not by itself establish independence, natural stimulus control, generalization, maintenance, causation, or a good client experience.
Define errorless teaching data categories
Independent accuracy uses correct responses before an added prompt divided by eligible opportunities. Prompted accuracy uses its own prompt-level denominator. Fidelity equals procedure steps completed as written divided by steps due. Later probe accuracy uses only the predeclared unprompted cohort.
Keep invalid events visible
Record missing materials, wrong cues, device outages, staff timing errors, safety stops, and withdrawal by reason. State the skill, natural cue, prompt, person, setting, version, and review period. Do not pool data after a material prompt or cue change without marking the new series.
Avoid universal benchmarks
The Leaf trial compared defined procedures for a narrow tact task with 28 autistic participants. The prompting review synthesizes varied studies and strategy variables. Neither supplies a universal independence percentage. The BACB outline and Ethics Code support integrity and evaluation.
Include the person's report
The ASHA AAC guidance supports continuous communication access. Record accessible reports of clarity, effort, comfort, and preference. A staff rating cannot replace the person's message.
A practical example
Across ten eligible opportunities, Hana responds independently four times, after a visual prompt five times, and pauses once. Staff deliver the prompt on time in 8 of 9 opportunities where it is due. Report 4/10 independence, 5/10 prompted response, 1/10 pause, and 8/9 prompt-timing fidelity.
Questions families can use
What is the denominator? Which prompt level applies? Are pauses and invalid trials separate? What do later unprompted probes show? Which version produced the data? What does the client report?
Sources
- Association for Behavior Analysis International, Basic Principles in Behavior Analysis
- Behavior Analyst Certification Board, BCBA Test Content Outline, 6th edition
- Leaf and colleagues, Comparing Error Correction to Errorless Learning: A Randomized Clinical Trial
- A Decision-Making Tool for Evaluating and Selecting Prompting Strategies
- Behavior Analyst Certification Board, Ethics Code for Behavior Analysts
- American Speech-Language-Hearing Association, Augmentative and Alternative Communication
Finni resources
