Training Framework

The DR Family: DRA, DRI, DRO, and DRL

Four procedures, one principle: reinforce what you want strategically and the problem behavior loses its job. The trick is picking the right one — because each targets the problem from a different angle, and the wrong choice means a confused dog and a frustrated owner.

The four differential reinforcement procedures compared

DRA — Reinforce a Specific Alternative

Differential reinforcement of alternative behavior means you reinforce a specific named replacement that serves the same function as the problem behavior. A dog that jumps on guests for attention is taught to sit for attention. When it sits, it gets petting and greeting. When it jumps, guests turn away.

DRA is the most versatile procedure because the trainer selects a behavior that is practical, teachable, and functional. Athens and Vollmer (2010) demonstrated that DRA works even without strict extinction of the problem behavior — when reinforcement quality, magnitude, and delay all favor the alternative, the largest effects emerge.

DRI — The Physical Guarantee

DRI is a subset of DRA where the replacement is physically incompatible with the problem. A dog lying on a mat cannot simultaneously jump on guests. A dog with four feet on the floor cannot counter-surf.

The strength is mechanical prevention. The weakness: physical incompatibility guarantees nothing about function. A dog trained to hold a down-stay during greetings through food reinforcement has not learned to request the attention that maintained the jumping.

DRO — Reinforce the Absence

DRO delivers a reinforcer when the target behavior hasn't occurred for a specified interval. The dog can do anything except the target behavior. Protopopova et al. (2016) demonstrated automated DRO for barking, reducing bark rates from 9.69 to 3.79 barks per minute and successfully thinning intervals to 1,200 seconds.

The limitation: DRO tells the dog what NOT to do but not what TO do. This can produce a confused dog cycling through random behaviors until it contacts reinforcement.

DRL — When Some Is Fine, Too Much Isn't

DRL reinforces the behavior when it occurs below a threshold rate. One alert bark is functional; sustained barking is not. The trainer reinforces the dog when it produces fewer than a criterion number of responses per interval. Becraft et al. (2018) found that signaled DRL — using discriminative stimuli to indicate when reinforcement is available — produces the most stable results.

Which DR Procedure Do You Need?

DRA when you have a clear, teachable replacement that serves the same function. DRI when you need a physical guarantee — especially for safety-critical problems. DRO when no single replacement is obvious. DRL when the behavior is acceptable at low levels. When in doubt, default to DRA — it teaches what TO do.

Always Pair Extinction with Reinforcement

When extinction is part of a treatment package that includes DRA or DRO, extinction bursts drop from 36% to approximately 12%, and aggression and other side effects decrease by more than half. Fisher et al. (2023) modeled this mathematically: dense DRA maintains overall reinforcement rates, preventing the reinforcement-rate decrease that drives bursts. The practical rule for dog trainers is unambiguous — always pair extinction with a reinforced alternative.

Every Dog Is Different

The right approach depends on why your dog does this — and that varies by temperament, history, and environment. The Synchrony coach can tailor these principles to your dog's specific behavior profile.

Sources: Vollmer & Iwata (1992), Cooper, Heron & Heward (2020), Athens & Vollmer (2010), Protopopova et al. (2016), Fisher et al. (2023). From the Data Dogs research briefs: ABA Methodology and Reinforcement in Practice.