Comparison of response prompting procedures in teaching numeral identification to autistic subjects.

Three autistic students learned to name numbers 1-10.

Two prompting methods were tested: constant time delay and system of least prompts.

Each student got both methods in alternating sessions.

Teachers measured how many sessions, trials, errors, and minutes each method needed.

Both methods worked for two students.

Constant time delay won on four out of five efficiency measures.

It needed fewer sessions, fewer trials, fewer errors, and less total time.

Only one student did slightly better with system of least prompts on errors.

Jones et al. (1992) looked at 36 studies and found the same pattern.

Their big review shows constant time delay keeps working across many skills and ages.

Eugenia Gras et al. (2003) used a different prompting style for compliance, but the core idea matches.

They also showed that how you fade prompts matters more than just reinforcing correct answers.

Carnett et al. (2020) taught questions instead of numbers, yet still used careful prompting.

All these papers tell the same story: plan your prompt fading, don't just reinforce.

If you're teaching any discrete skill to autistic learners, start with constant time delay.

You'll save hours of therapy time and reduce student frustration.

Pick a 3-second delay, stick to it, and watch the data.