Preliminary comparison of two negative reinforcement schedules to reduce self-injury.

The team tested two ways to stop self-injury that is kept going by escape. Both ways let the child take a short break, but only after different things happened.

In DNRA the child had to ask for a break with a card or sign. In DNRO the child only had to stay safe for 30 seconds. The two rules swapped every session so each kid served as their own control.

Both plans cut self-injury to near zero and helped the kids finish more tasks. DNRA, the plan that required asking, worked faster and kept self-injury lower.

Kids also asked for breaks more often under DNRA, showing they learned a real replacement skill.

Lovaas et al. (1969) first showed that turning off attention can stop self-injury. Einfeld et al. (1995) moved past that by asking, 'What should the kid do instead?' Their answer: teach a clear escape request.

Wanchisen et al. (1989) tried a close cousin, DRI plus interruption, and saw mixed results. The 1995 study shows that skipping interruption and using DNRA gives cleaner, faster gains.

Miller et al. (2022) and Boyle et al. (2021) pick up the story later: once the child can request escape, you can stretch the work time and still keep behavior low. The 1995 paper gives the first step—install the request—before you thin the schedule.

If your functional analysis says escape, don’t just block the hits. Pick one clear way the learner can say 'I need a break' and reinforce that request every time at first. DNRA beats DNRO for speed and durability, and it sets you up for safe schedule thinning later.