Stimulus (instructional) fading during extinction of self-injurious escape behavior.

Three kids with developmental delay hit themselves to get out of work. The team kept extinction in place—no more escape—but also faded the work. They started with zero demands, then added one, then two, until the kids did their old load again.

The researchers tracked self-injury minute by minute. They used a multiple-baseline design across the three kids.

Self-injury dropped to near zero for all three kids within the first week. Gains held when demands returned to the original level.

Parents and teachers said the kids stayed calmer and still finished their tasks.

Slocum et al. (2025) ran a bigger test and found the same thing: fading demands beats plain extinction for speed. Their 2025 data now supersedes this 1993 small study.

Carr et al. (2003) used a different fade—mixing easy and hard tasks—but also got zero escape SIB. The tactic differs, yet the principle matches.

Silva et al. (2025) took the idea further. After extinction they faded the room cues, not the work, and cut relapse risk. Same combo, new problem.

If a learner hurts herself to escape tasks, start with almost no demands while you hold extinction. Then raise the work in baby steps. This 1993 paper proved it works, and newer RCTs say it still beats extinction alone. You can apply it Monday: cut the task load in half, reinforce compliance, and add one new demand every two days.