Manipulation of potential punishment parameters in the treatment of self-injury.
Punish the first twitch of self-injury and you can stop the whole blow, but newer work shows you can do it without pain.
01Research in Context
What this study did
One boy with severe intellectual disability hit his head 300-600 times per hour.
The team first tried mild punishers: a brief hair tug, then a short shock. Each step cut the hits but did not stop them.
Next they added continuous arm splints and a final rule: any early sign of hitting brought one quick shock. After that, hits stayed at zero for almost three years, even in new rooms and with new staff.
What they found
Linking the shock to the very first move of the self-injury chain wiped the full chain out.
The boy kept his hands down without restraints once the early cue became the signal for punishment.
How this fits with other research
Stokes et al. (1980) got the same zero-hits result with just a cool water mist to the face. Their mild punisher matched the 1975 shock study in direction and size, showing you can swap pain for water and still win.
Rayfield et al. (1982) took the 1975 arm-splint idea and made it contingent instead of always-on. They added rewards for safe hands and then faded the gear. SIB stayed low, proving you can keep the protection but lose the restraints.
van der Miesen et al. (2024) pooled 129 newer SIB studies and found parent-run home plans work as well as clinic shocks. The meta-analysis includes the 1975 case, but shows today’s teams reach the same zero with praise and toys, not electricity.
Kohler et al. (1985) tested positive-practice overcorrection. Like the 1975 study, it only worked when tied to the act. Both papers agree: timing beats intensity.
Why it matters
You now know that catching the first millisecond of a self-injury chain can erase the whole chain. If you still use punishment, pair it with early warning signs and plan to fade it fast. Better yet, copy the 1980 water-mist or 2022 caregiver-playbook routes and skip the shock box entirely. Either way, measure the early cue and hit it first.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Film one session, list the earliest body movement before each hit, and program a brief consequence right then—start with a mist, not a shock.
02At a glance
03Original abstract
The self‐injurious behavior (SIB) of a blind, profoundly retarded male was observed within a field setting as multiple forms of punishment were applied. The effects of hair‐tug punishment were first systematically examined, then hair‐tug and electric shock punishment were administered in varing ratios (of punishment deliveries per SIB response) supplemented by continuous restraint contingencies and made contingent on recurrent antecedents of SIB cycles. Treatments were applied over six months of consecutive daily sessions ranging from 30 min to 8 hr duration. Posttreatment followup observations were taken 12, 24, and 34 months after treatment. Partial suppression was produced as a direct function of hair‐tug punishment in the initial phase. A generally progressive decrement was produced on already lowered SIB rates with the succeeding components of the treatment package. Extended periods of total SIB suppression occurred as punishment was rendered contingent on antecedent components of each SIB response cycle. Followup observations of nearly 3 yr duration indicated total suppression generalized over all settings in which the subject functioned. The findings were related to the contrasting limits prevailing in most reports of punishment‐based SIB suppression.
Journal of applied behavior analysis, 1975 · doi:10.1901/jaba.1975.8-458