Effects of warning stimuli for reinforcer withdrawal and task onset on self-injury.

Researchers tested a simple warning package on kids with autism who hit or bit themselves. Before taking away a toy or starting work, staff showed a picture card and said 'Time to work' 30 seconds early.

They kept two safety tools running at the same time: extinction (no escape from the task) and noncontingent reinforcement (toys or snacks given for free). A two-minute delay without the warning served as the control.

Self-injury dropped to low, safe levels when the warning package was used. During the two-minute delay control, SIB stayed high.

The warning let the kids 'brace' for the change. Combining it with extinction and free reinforcers kept problem behavior down without any extra punishment.

Castañe et al. (1993) did something similar five years earlier. They added instructional fading to extinction and also saw fast SIB drops. Both studies show that giving clients a 'heads-up' before a hard moment makes extinction safer.

Hatton et al. (1999) looked at 41 cases where extinction was used alone. Almost half exploded into bursts or aggression. The current study avoided that trouble by wrapping extinction inside NCR and a warning cue — a live example of the buffering C et al. called for.

Tereshko et al. (2017) extended the idea two decades later. They paired escape extinction with protective helmets and saw durable SIB reduction. Together, the three papers form a timeline: add antecedent cues (1993, 1998), then add safety gear (2017), all while keeping extinction in the mix.

If you run extinction for escape-maintained SIB, drop the 'surprise' factor. Show a picture and give a short verbal cue 30 seconds before the transition. Keep noncontingent reinforcers flowing and do not let the child escape the task. This cheap package cuts bursts and keeps rates low from day one.