What Treat is Your Favorite?

---

---

Most of us pay close attention to the types and flavors of treats that our dogs enjoy. This is of interest because many trainers and owners use food treats as a primary positive reinforcer and often rank treats from those that appear to be relished (high value) to those that are classified as acceptable but not particularly yummy (low value). The way that we have traditionally determined these values has been subjective at best – we watch how our dog responds to a given treat, how well the treat performs as a reinforcer, and then we “guestimate” to classify treats as preferred vs. not preferred by that particular dog.

However, a lot of factors can influence subjective treat valuations. These include how motivated our dog is at that moment, the type or difficulty of the behavior we are targeting, the number of treats the dog has already consumed, and our own biases regarding what dogs like or dislike.

Dog food preferences are of great interest to nutrition researchers as well. I wrote an essay about a study conducted at the University of Illinois that used stuffed Kong toys to test food and ingredient preferences in dogs. That test allowed multiple selections and worked quite well as a measure of food and ingredient preferences (see Taste Tests). While fun to conduct, the test in that study did require some equipment (multiple identical food-delivery toys), and generally would be difficult for a trainer to conduct without having another human helper available.

Recently, Drs. Mindy Waite of Carroll University and Tiffany Kodak of Marquette University conducted a set of experiments that tested the usefulness of a treat preference test that required nothing more than a set of treats, one trainer, and of course, a hungry dog who was eager to show her preferences. And, this being science, they not only created the test – they studied its efficacy with a group of dogs and then validated the results! This research was generously funded by Fear Free. Let’s take a look at what the researchers did:

The Study

A group of 11 healthy adult dogs and their owners participated. Training took place in the owner’s home and training sessions were videotaped for subsequent analysis. The study was conducted in three phases:

• Teaching Discrimination: The first step involved teaching dogs that they have a choice (i.e. to discriminate). This was conducted by pairing two treats; the first was always a treat that the owner knew to be low value or disliked (for example, Stanley, my Toller, believes that bananas are the devil’s food) .

---

• The second was a treat that was known to be relished by the dog. The owner held one treat in each hand, with a small bit exposed. The dog was allowed to first sniff and lick each treat alternatively; the owner then raised the hand to chest level. Both hands were then presented simultaneously and the dog was allowed to choose (and eat) the preferred treat (Note: Detailed instructions are included in the paper). The researchers took care to prevent side bias (the dog choosing a preferred hand) and owner bias (the owner unintentionally influencing their dog; aka Clever Hans Effect). After multiple repetitions, the dogs were tested to ensure that they were showing discrimination – in other words, the dogs understood that each hand held different treats from which they could choose.

• Treat Preference Testing: Now the fun begins! This type of test is called a “paired-stimulus test,” which means that the dog is offered two choices in each test, one type of treat in each hand. Multiple food treats can be tested, paired in a randomized order. The procedure is identical to the discrimination training, with the exception that the dog is now presented with treats that might be relatively close in value. The number of trials that are needed depends completely on the number of treats that the owner is testing. Below is a video that Dr. Waite provided demonstrating the testing procedure (video used with permission and our gratitude; Dr. Waite and her dog Gus).

---

• Validation of the Test: For those of you versed in behaviorism, this part involved what is called “Progressive Reinforcement” (PR). In this test, a progressively increasing number of indications by the dog (in this case, a touch response) is needed prior to the dog actually receiving the selected treat. The end point is extinction of the indicating behavior – the dog stops trying. The intent is to objectively measure how valuable the treat truly is to the dog, with the premise that dogs will try longer (more reps) for higher value treats. (Note: Personally, I avoid using extinction in dog training, because any training approach that employs extinction is inherently frustrating for the dog. I make this note here because I would not do this follow-up validation test with my own dogs. Everyone of course, can make their own decision about this. Note also that this step is not necessary in practice. Rather it was an important validation step in the study).

Results

Not only was this test easy and fun for dog owners to perform, it accurately helped them to identify treats that their individual dog loved best! Here is a summary of the study results:

• Dogs Showed What They Like: Across the 11 dogs, cheese was the most highly preferred treat for 5 dogs (45 %) and kibble was least (or almost least) preferred by 8 dogs. Because owners selected the treats that they tested with their dog, the type of kibble (or cheese) was not consistent. What was consistent however, was that dogs showed distinct and highly individual preferences using this paired test.
• Preferences were Validated: When tested using the PR protocol, dogs showed a higher “break point” (i.e. extinction of the touch behavior) with treats that they had identified previously as high value compared with those that they did not prefer. In other words, dogs continued to “touch” for longer in the face of no +R when a high-value treat was used compared with when a low-value treat was used.
• Owner Compliance and Ease of Use: A set of tests that measured owner compliance and consistency found that the trainers easily completed the test with their dogs. When surveyed, owners perceived little to no negative outcomes for their dogs.
• Smell AND Taste: An important addition to the procedure used by these researchers (compared with earlier work) was that dogs were allowed to both sniff and lick (taste) the food when it was initially presented to them. The researchers speculated that the short “taste” that licking provided aided dogs in making their preferences known.
• Owner Predictions (versus Dog Preferences): Owners of 10 dogs made a set of predictions, prior to starting the training, regarding which treats they believed that their dog would prefer. Of these predictions, only two were close to matching their dogs demonstrated preferences, while eight dogs diverged (sometimes widely) from their owners’ predictions. (This was a surprising result!)

---

---

Take Away for Dog Folks

There are SO many things to love about this study (including of course, that it was an in-home study, something that we strongly promote at The Science Dog)!

This paper provides dog folks with a validated treat-preference test that is easy-to-use and enjoyable for dogs and their people! Because the researchers focused on a paired-stimulus version that does not require extensive pre-training or extra equipment, trainers can use this test quickly and repeatedly throughout their dog’s life to assess current likes and “less likes” for training treats. Interestingly, a second study, published by a different group of researchers, found that while both paired-choice and multiple-choice tests were successful with dogs, using the paired-choice test was more efficient (2). Corroboration in science is so lovely! Finally, including the opportunity for dogs to taste each treat first via licking was a brilliant addition that, according to the paper, may be helpful for obtaining accurate assessments.

---

Interested in learning which treats your dog really likes best? Give this new preference test a try (and let us know how it goes in the comments section below)!

---

Cited Papers:

1. Waite MR and Kodak TM. Owner-implemented paired-stimulus food preference assessments for companion dogs. Journal of Experimental Analysys of Behavior 2023; 1-16. DOI: 10.1002/jeab.846.
2. Payne SW, Fulgencio CT and Aniga RN. A comparison of paired- and multiple-stimulus-without-replacement preference tests to identify reinforcers for dog behavior. Journal of Experimental Analysis of Behavior 2023; 1-13. DOI: 10.1002/jeab.857.

---

NEW! Free Science Dog Webinar

Learn how the term “Complete and Balanced Nutrition” first found its way onto pet food labels, what it (actually) signifies, and what it does not……. (and how to avoid confusion between the two).

Join us for this complimentary offering!

---

Take a Look at this Popular Science Dog Courses Webinar!

The Scoop on Poop

An evidence-based webinar examining how diet and life style factors influence fecal quality and GI health in dogs. This comprehensive discussion examines ALL of the factors that go into optimal fecal quality (and quantity) in dogs and what owners can do to support it.

Designed for dog owners,  trainers, veterinary health professionals, and  pet food industry professionals. View the webinar on-demand, plus enjoy the opportunity to chat with the presenter and ask questions.

---

Also From The Science Dog Courses

Earn your Masters Class Certificate in Canine and Feline Nutrition by completing four CORE courses! (Basics, Puppies to Seniors, Dog Food Smarts and Feeding Felines)

Masters Class Certificate in Canine and Feline Nutrition

A comprehensive program in basic nutrition, feeding management throughout life, & food evaluation and selection, designed for dog and cat enthusiasts & professionals.

Learn the latest research regarding canine and feline nutritional science, recommendations for food assessment & selection, and best feeding practices for dogs and cats throughout life

---

Not sure where to start?

Learn the Basics

---

---

---