Short answer: Research comparing force-free vs aversive training points the same way: reward-based methods work at least as well, and aversive methods are linked to more stress and behavior problems. The American Veterinary Society of Animal Behavior (AVSAB) states that there is no evidence aversive training is necessary. The studies have limits, but none of the key sources below found an advantage for aversive methods.

What do "force-free" and "aversive" training mean?

Force-free, or reward-based, training teaches a dog what to do by adding something they like, such as food, play or praise, when they get it right. Aversive training relies on something unpleasant to stop or drive behavior, such as leash jerks, physical corrections, prong collars or electronic shock collars. AVSAB describes aversive methods as those that "rely on application of force, pain, or emotional or physical discomfort."

What do the key studies and position statements say?

What did each source find?

AVSAB (2021). The position statement on humane dog training says reward-based methods should be used for teaching common skills and addressing unwanted behavior, and that aversive methods "should not be used in canine training or for the treatment of behavioral disorders."

Hiby, Rooney and Bradshaw (2004). In this survey of 364 owners, owners' ratings of obedience went up with the number of tasks trained using rewards, but not punishment. Reported problem behaviors went up with the number of tasks trained using punishment. The authors concluded punishment "may represent a welfare concern without concurrent benefits in obedience."

Ziv (2017). This review of 17 studies concluded that people working with dogs "should rely on positive reinforcement methods" and avoid positive punishment and negative reinforcement as much as possible. A free summary is available from Purdue's Canine Welfare Science center.

Vieira de Castro and colleagues (2020). This study in PLOS ONE concluded that aversive-based methods, "especially if used in high proportions, compromise the welfare of companion dogs both within and outside the training context."

China, Mills and Cooper (2020). In this trial in Frontiers in Veterinary Science, reward-focused training was more effective for recall and sit than training with e-collars. The authors wrote that the findings "refute the suggestion that training with an E-collar is either more efficient or results in less disobedience."

What are the limits of this research?

Many studies are surveys or observational, so they show links rather than proving cause. Samples are often small, and dogs aren't always randomly assigned to methods. The Ziv review flags these issues itself. Even so, the sources point in the same direction, and none of them found aversive methods necessary.

How do I put this into practice, step by step?

  1. Choose reward-based training for everyday skills and for problems like reactivity.
  2. Ask trainers what happens when a dog gets it wrong. Look for answers about changing the setup, not corrections.
  3. Skip prong, choke and shock collars. Pick equipment that doesn't rely on pain or discomfort.
  4. Use management to prevent problems while you train.
  5. Talk to your vet if behavior changes suddenly or your dog may be in pain.
  6. For serious issues such as aggression or anxiety, look for a qualified, reward-based behavior professional or a veterinary behaviorist.

To put reward-based training into practice, The Handler guide covers timing, rewards and reading your dog step by step. The first chapter is free to read.

Sources

This article is general education, not veterinary advice. If your dog seems unwell, is in pain, or something changed suddenly, call your vet. If it could be an emergency, don't wait.

Frequently asked questions

Is force-free training as effective as aversive training?

The sources here that compared training results found reward-based methods worked at least as well. China and colleagues (2020) found it more effective for recall and sit than e-collar training.

Does aversive training harm dogs?

The research links aversive methods with more stress-related behavior and more reported behavior problems. AVSAB recommends against using them.

What does AVSAB recommend?

AVSAB states that reward-based methods should be used for all canine training and that aversive methods should not be used in training or for treating behavioral disorders.

Are e-collars ever necessary?

In the 2020 trial by China and colleagues, the authors found no evidence that e-collar training is necessary, even for recall problems.

What are the weaknesses of this research?

Many studies are surveys or observational, samples are often small, and dogs aren't always randomly assigned. The findings still point the same way.

Explore this topic

Dog Pulls on the Leash? How to Teach Loose-Leash Walking