Hi Reader,

Counterconditioning is one of the most frequently recommended techniques in dog behaviour modification, and most of us use it (or some version of it) almost daily. So when a systematic review on counterconditioning in dogs came out in 2024, I was very cursious.

Some of the findings were quite uncomfortable, in a useful way.

The authors (Shnookal, Tepper, Howell and Bennett) looked at every available study testing counterconditioning in companion dogs. They found 14 studies in total, 12 peer-reviewed papers and 2 dissertations. That’s a small number for a technique we recommend constantly, and the studies also varied a lot in how they were designed and what they tested.

Still, some clear patterns came out.

Two types of counterconditioning

The first thing the review showed is that the term “counterconditioning” is actually being used to describe two different procedures.

In some studies, the trigger is paired with food regardless of what the dog is doing in that moment. The visitor walks in and the food always follows. The thunderclap sounds and the dog always gets the treat, regardless of the behaviour. This is classical counterconditioning, and you probably use this in your daily practice.

In other studies, the food is paired with a specific behaviour the dog has been trained to perform, like a sit, a down, a place cue, or a focused look at the trigger. The food is contingent on the behaviour, which is technically operant counterconditioning.

Many studies running this procedure still called it counterconditioning anyway. So when we talk about whether counterconditioning works, we’re actually talking about two different things working through two different learning systems.

What worked and what didn’t

For aggression toward visitors and other dogs, counterconditioning worked quite well, but specifically the operant version where the food was contingent on a behaviour like a bed-stay or a focused look.

For noise phobias and undesirable kennel behaviours, the classical version worked better.

But for separation-related behaviours, counterconditioning wasn’t effective at all.

The longest duration any dog reached alone without showing problem behaviour, after weeks of expert-supported work, was 5 minutes. So separation distress is clearly different from the other undesirable behaviours.

The reviewers suggest that this is because counterconditioning is most effective when we pair a trigger with a reinforcer, which means the technique needs something that arrives for the dog to associate with the food.

With separation, there’s nothing arriving for the food to associate with. The owner is leaving after all. And with the owner leaving, the support system of the dog leaves too. Which means the issue is really the inability to regulate their emotions when alone.

The big limitation of counterconditioning research

One of the main limitations of the studies in this review is that they did not include any physiological measures, such as heart rate variability, salivary or hair cortisol, or other indicators of autonomic activation.

So everything we know about whether counterconditioning “works” comes from observable behaviour, which is only part of the picture.

The reviewers themselves point this out. They write that it’s possible dogs learn to display the desirable behaviour to obtain reinforcement, without changing the underlying emotion associated with their original behaviour, particularly in operant counterconditioning interventions.

So the dog who lies on his bed when the visitor arrives might still be in full sympathetic activation. The behaviour stopped, but we don’t actually know whether the dog is feeling any better.

We have to be careful not to confuse trained compliance with actual safety and emotional regulation.

I wrote a full breakdown of the review on the blog, with all the studies broken down by behaviour category and a section on best practices when using counterconditioning.

Read it HERE.

Source: Shnookal, J., Tepper, D., Howell, T., & Bennett, P. (2024). Counterconditioning-based interventions for companion dog behavioural modification: A systematic review. Applied Animal Behaviour Science, 276, 106305.

Cheers,

Dr. Else

​​​​​