Wondering how often you should train your dog? Ask Science! (Part 1)

When I first began training dogs, all of the books and trainers advised me to diligently train my dog every day to produce the best results. When I asked how long I should train my dog, I was often answered in units of time (e.g., one hour or thirty minutes) while the more force-based trainers advised, “until the dog gets it right.” Armed with this advice I logically concluded that I should spend an hour (or more!) each day training my dog. Thankfully, my knowledge of how dogs learn and dog training  has grown since those initial summer evenings spent tirelessly working my poor 4-H dog. Now, as I train my current dogs and advise others on the most optimal way to train their dogs, I have science in my corner!

What they did: In 2008 researchers looked into the relationship between the number of training sessions per week and learning in 18 Beagles. One group of Beagles were trained once a week with 6 – 8 days between sessions ( Sound familiar? This is how most weekly dog classes are set up), whereas the other group of Beagles were trained five times per week with approximately 24 hours (72 on weekends) between training sessions. The researchers wanted to test what effect massed training (training sessions five times a week) and spaced training (training sessions once a week) had on learning.

Every dog was shaped via positive reinforcement to touch a mouse pad with their front paw. (This might seem odd but that is exactly what the researchers wanted, it was important for the dogs to not have had any experience performing a similar exercise.) Of note, is that the researchers defined a “training session” by an exact number of trials and the length of time between one trial and the next (intertrial interval) which is different than how (most) dog trainers define a training session, in that, most dog trainers say a session is 30 minutes/1 hour, etc. For anyone who is interested, there were a total of 15 trials each session.

What they found: Contrary to what those old dog trainers told me back when I was first starting out, training once per week appears to be beneficial. Dogs in the group trained once per week learned the paw exercise in fewer sessions and had higher success rates during all phases of the training than the dogs trained five times a week. Overall, it appears that at least when learning a specific task, weekly training sessions result in better learning.

What this means for us: Those old 4-H trainers were wrong! Although generalizing too much from one study is always risky, further research (which I will cover in part 2) supports what these researchers found. So, when you are training a new behavior, it might be prudent to scale back on the number of training sessions you do with your dog. It also appears that the standard class schedule for most dog classes (once per week) is supported by science! Take that to your next class!

Note: This study pertained to training new behaviors and did not cover anything about maintaining behaviors or working to make behaviors become automatic. Many people in the performance sports world engage in mass training in the hopes of having their dogs’ behaviors and reactions become automatic in certain contexts. There is no research in this area, to my knowledge.

I would like to issue an apology to my  sweet American Cocker Spaniel/Beagle mix, Kes for all of those long, arduous evenings spent training during Nebraskan summers. Kes is pictured on the featured image of this article.

References :

Meyer, I., Ladewig, J., (2008). The relationship between number of training sessions per week and learning in dogs. Applied Animal Behaviour Science, 111. 311 – 320. doi: 10.1016/j.applanim.2007.06.016

Looking into What Your Dog is Looking at.

Have you ever wondered what part of your face your dog pays attention to the most? Good news! There’s science for that!

Researchers at the University of Helsinki are looking into what might be going on with dogs when they look at the faces of other dogs and humans. In particular, the researchers were interested in whether or not dogs pay more attention to a dog or human face that is threatening (i.e., aggressive dog face or angry human face). We humans (as well as other primates) pay more attention to things that could potentially be threatening and after all, this makes sense from a evolutionary standpoint. No species ever got anywhere ignoring what could mean bad news. Now, good ol’ fashioned eye-tracking technology is showing us how are dogs do this!

In research involving animals (or humans too young to speak), the go-to way to figure out what grabs and holds attention is eye-tracking software. In this case, researchers utilized eye-tracking technology to track what dogs focus on the most when they look at faces by analyzing the length of time the dog looked for and the order of where the dog looked first. This may not sound like a big deal, but whatever a dog gazes at the longest, is what is capturing the dog’s attention the most, and by extension, what is most important to the dog at that moment in time. By analyzing data from multiple dogs’ gaze patterns, researchers can begin to get an idea of just what your dog is paying attention to when he is staring at you.

How they did it. For this study, researchers clicker-trained 31 domestic dogs to lie in front of a blank monitor while resting their head on a chin rest. The dogs were trained do this unprompted, were not encouraged to look at the monitor, and held the position for at least 30 seconds. Researchers then projected images of canine and human faces onto the monitor in the hopes of figuring out whether the emotion of the faces affected what facial features the dogs looked at the most.

Three separate emotional categories were used: Threatening (aggressive dog faces and angry human faces), Pleasant (Happy dog faces and happy human faces), and Neutral (expressionless dog faces and expressionless human faces). Each face was presented to an individual dog with a break between the presentation of one face and the next.1 The order of species and emotion of the faces was randomized so the dogs would not be able to anticipate what might be presented next.

What they found: Dogs look at faces using the same strategies that we do. Dogs adjust how they look at faces based on the emotional content of the face they are looking at. Regardless of species, dogs fixate faster on the eyes and midface area than on the mouth, and the eyes are more likely to be what the dog looks at first. This might be because the eyes are such an important area for communicative intent in both dogs and people. Where things begin to differ between dogs and humans, is the mouth.

When viewing pleasant human faces, humans look at the mouth presumably because the smile is the biggest indicator of good mood in our species. Dogs on the hand, look longer at the mouth of threatening and neutral dog faces. This makes sense when you think about how important a dog’s mouth is in conveying aggression (i.e., wrinkled muzzles, snarling, bared teeth, tight lips).

What this means: Dogs make use of the emotional content on your face (and the faces of other dogs). This, as far as we know, is the first evidence that a non-primate uses emotional information when looking at faces! Which, you know, is pretty awesome.

  1. A break between the presentation of each face (what scientists call an inter-trial-interval) is needed so that the presentation of one face is not influencing the presentation of the next face, and so that there is a clear start and stop point for data analysis.

Tl:dr: dogs adjust how and where they look at the faces of other dogs and humans based on the emotion those faces are displaying.

References

Somppi, S., Tornqvist, H., Kujala, M. V., Hanninen, L., Krause, C. M. & Vainio, O. (2016). Dogs evaluate threatening facial expressions by their biological validity-evidence from gazing patterns. Plos One, DOI: 10.1371/journal.pone.0143047