Reducing pawing in horses using positive reinforcement.

Austin et al. (2015) worked with three horses that pawed the ground.

The team used a DRO schedule. If the horse kept its hoof still for a set time, it got a carrot.

They started with short waits and slowly made the waits longer.

Pawing dropped for all three horses.

One horse needed tastier treats to stay quiet. The others worked for plain pellets.

The study shows DRO can replace punishment when you pick the right reward.

Mantzoros et al. (2023) used a cousin schedule, DRL, to cut vocal stereotypy in teens with autism. Both studies prove you can thin the schedule and still keep the gain.

Einfeld et al. (1995) compared two negative-reinforcement flavors for self-injury. Like E et al., they showed the schedule you pick matters, but they worked with escape behavior, not attention.

Dugdale et al. (2000) pitted NCR plus a tiny delay against DRO and hit near-zero problem behavior. Their NCR gave more treats with less work, yet E et al. still got good results with simple DRO. The gap is species and setting, not principle.

You can cut pawing, vocal stereotypy, or pica without punishment. Pick a differential-reinforcement schedule, test reinforcer value, and thin slowly. Watch the client, not the cookbook.