Increasing the independent responding of autistic children with unpredictable supervision.

Three autistic children worked at a table while the therapist left the room on a schedule. The schedule was random: the adult might pop back in after 30 seconds or after five minutes.

The team compared this random check-in plan to a fixed plan where the adult returned every two minutes. They measured how long each child stayed on task while the adult was gone.

All three kids stayed on task longer when the adult used the random schedule. The surprise visits kept the work going better than the clockwork every-two-minute check.

Aman et al. (1987) took the same idea into grocery stores and parks. Their autistic clients kept good community behavior even when the therapist rewarded them only now and then.

El-Boghdedy et al. (2023) and Jessel et al. (2017) later added gradual fading and momentary DRO. They showed teens can stay on task too, but they slowly thin both prompts and prizes.

Leon et al. (2023) flips the coin: unpredictability in activity switches made kids aggressive. The difference is domain—random check-ins help, random transitions hurt.

You can step away without losing control. Mix up your return times instead of setting a watch. Start with short gaps, then stretch them once the child shows steady work. Pair the plan with praise or tokens when you do return. This keeps the value of “adult eyes” alive even when you are out of sight.