A variable influencing the performance of generalized imitative behaviors.

Peterson et al. (1971) worked with 12 preschool kids who had no disabilities. The team first taught each child to copy 10 simple actions like clapping or touching their head.

Next they tested whether the kids would copy NEW actions. Sometimes the teacher stayed in the room. Sometimes the teacher left. They counted how often kids copied without any reward.

When the teacher stayed in the room, kids copied new actions about a large share of the time. When the teacher left, copying dropped to a large share.

Just having an adult watch made generalized imitation jump up four-fold. No extra praise or candy was needed.

Pisacreta (1982) showed pigeons also need the trainer present to peck a moving key. Both studies prove that 'who's watching' can control newly learned skills.

Lazar (1977) found that special sound cues helped pigeons learn lever chains faster. Together with F et al., this tells us both people and signals can act like hidden rewards.

Williams et al. (2002) taught pigeons with only 2-second picture flashes. F et al. used short trials too. Both show quick lab sessions can still create strong stimulus control.

Before you run a generalization probe, check the room. If the usual adult steps out, a child may suddenly 'forget' skills they actually have. Keep the familiar person in view during first tests of new imitation, then fade yourself out once the skill is solid.