The relationship of self-stimulation to learning in autistic children.

Fixsen et al. (1972) worked with three autistic children who rocked, flapped, or spun objects during lessons. The team ran teaching trials for a simple picture discrimination. On some days they let the kids self-stim. On other days they stopped the self-stim with gentle hand prompts and brief restraint.

No shocks, no scolding—just blocking. The goal was to see if less stereotypy helped the kids learn the task.

When self-stimulation was blocked, all three children quickly learned to pick the correct picture. When self-stim was allowed, learning flat-lined even after many trials.

The takeaway: high stereotypy during teaching blocks new learning.

Barthelemy et al. (1989) later got the same drop in stereotypy without touching the child—just by clearing clutter and adding clear work stations. Their classroom redesign gives you a non-aversive way to copy L et al.'s effect.

Tiger et al. (2017) conceptually replicated the finding by teaching kids to stop stereotypy on cue with S– stimuli. Once the cue controlled the behavior, learning tasks got easier, echoing the 1972 result.

Ahrens et al. (2011) seems to disagree. They watched autistic children in free play and saw that stereotypy bouts often ended on their own. But the settings differ: free play vs structured teaching. During instruction you can't wait for the behavior to stop—you need the learner ready now.

If a child is rocking or flapping during your trials, don't keep drilling. First bring stereotypy down—by blocking, by arranging the room, or by teaching an inhibitory cue—then run the teaching trials. Learning will speed up and you will save precious session time.