Food for Thought: Hungry Dogs and Training
Almost all trainers think that when you’re using food rewards, it’s best to start with a hungry dog. But a new study suggests that isn’t always true. Sometimes, it might be best to give your dog a good breakfast before he starts work.
Like every other dog trainer on the planet who uses , I tell my clients that they can teach most effectively if their dog is a little hungry. Train right before mealtimes, I say, or go ahead and use your dog’s whole meal, in bits and pieces, for food rewards. Scientists conducting experiments to study learning often keep animals really hungry, feeding them only enough at meals to maintain 85 percent of their free-feeding body weight. (This has always seemed unkind to me, by the way. I would never recommend it – I’m just describing the assumptions many of us make.)
Sponsor: This episode is brought to you by GoDaddy.com. GoDaddy.com offers everything you need to make a name for yourself on the Web, including domain name registration, website hosting, and more. Get 20% off your hosting plan purchase at by using the code hostpod51 at checkout.
But what if we were wrong? What if a well-fed animal actually performs better?
A Good Breakfast Helps Humans Think Better …
The researchers Holly Miller and Charlotte Bender realized that it seemed strange for animal trainers to insist that our students learn better when hungry. After all, we human beings are also animals – of the species Homo sapiens – and the opposite is true for us. When human children get breakfast, they do better on tests of memory. They’re better able to pay attention. Their problem solving skills improve. And all of that makes sense, because the brain is a huge consumer of energy, in the form of glucose. Which we get from digesting food. If hunger gets in the way of humans’ brainpower, why wouldn’t it get in the way of our dogs’ brainpower, too?
… So Why Do We Assume That Animals Are Best Trained When Hungry?
So Miller and Bender designed an experiment to test our assumption that animals are best trained when hungry. (Miller, Holly C., and Charlotte Bender, “: Dogs (Canis familiaris) Search More Accurately When They Are Less Hungry,” Behavioural Processes 91 [201, 313-17.) They divided their test dogs – these were all pets whose guardians had volunteered them – into two groups. All the dogs were fasted overnight, and then one group got tested 30 min after having either breakfast or nothing, whereas the other group were tested 90 min later. Both groups were invited to find a treat that they had seen the experimenters hide in one of several containers. Each of those containers also held a bit of chopped Oscar Mayer wiener that masked the smell of the treat. That way, the dogs would have to remember where the treat had been hidden – they couldn’t just sniff it out.
All of the dogs in each group were tested twice, once after breakfast and once while still hungry, so that performance could be compared for each dog. The dogs were tested one at a time, by the way. You don’t have to picture 14 dogs, 7 of whom are famished, homing in simultaneously on the same treat.
Dogs Did Best with Breakfast!
When did the dogs do best at finding the hidden treat on the first try? Drumroll: When they were tested 30 minutes after eating breakfast. That was when their memory for the treat’s location was best. At 90 minutes, there was almost no difference between hungry dogs and breakfast dogs – that breakfast benefit was gone.
Unfortunately, the demonstration that a good breakfast improves dogs’ cognitive skills is somewhat clouded. That’s because the researchers had their dogs do a 10-minute sit-stay before the search. The self-control of a sit-stay involves what are called executive processes, a cognitive function that (like everything else about the brain) vacuums up available glucose to use as fuel.
Another Possible Explanation?
Also, although Miller and Bender believe that the increased blood glucose provided by a breakfast accounts for their results, they explain that there’s another possibility: the hungry dogs may have beenso eager to find the food that they made mistakes out of sheer excitement. The same possibility applies to the dogs who searched 90 minutes after being fed – maybe they also were hungry enough to make them overeager. (This is a recognized pattern in learning, called the . The very short version is that eagerness improves performance, but only up to a point. Overshoot the eagerness sweet spot, and performance falls off again.)
As for me, I wondered whether some of the dogs who went for one of the “wrong” containers actually just said to themselves, “The heck with that wiener I saw the guy hide – I smell deli meat in every last one of these babies.”
But Miller and Bender’s results are, well, food for thought. They’re consistent with other studies of how eating and hunger affect the hard work of thinking and learning. They even match my experience with my own dogs, all of whom have been just as focused on training after dinner as before. I’ve always believed that this was because reward-based training makes the work of learning pleasurable, beyond just the satisfaction of physical hunger. But maybe having a good meal on board was actually helping my dogs stick with work.
Sometimes, a Little Hunger Might Still Be Best
On the other hand, there might be times when it’s more important for a dog to really, really want the food you’ve got than it is for her to maximize her brainpower. I train in New York City, where we usually can’t control a dog’s environment for purposes of behavior modification. In that context, it’s often crucial to be able to just plain distract a dog from something she might be frightened of or aggress toward. A little bit of hunger might be to our advantage then.
I’ve also found it useful to set aside at least some of a newly adopted dog’s daily food for . The idea is that the more often I can catch the dog and reward her for doing something I like, whether that’s choosing to lie down on her bed or keeping quiet when the doorbell rings, the more quickly she can learn the rules of her new home. (And also learn how to go about getting what she wants in life, as I described in a recent episode about .) How can I balance eagerness for food against the need to keep a thinking animal’s brain consistently fueled? Miller and Bender suggest that working dogs might benefit from snacking over the course of the day, so maybe my approach doesn’t need too much tweaking after all. Phew.
One other point. The dogs in the breakfast study were all known to be motivated by food rewards. Well, most dogs do like to eat – otherwise I wouldn’t hear so many human complaints about . But some get less worked up about food than others. For dogs who need a little more motivation to work for food, training before meals might be best even if there’s a tradeoff in cognitive strength. (You can use non-food rewards such as play, of course, but nothing beats food for the number of reps you can get in the training time available.)
There’s a lot to think about here. (I think I’d better have a biscuit.) But I’m pretty sure I just got a little less quick to suggest my clients train when their dogs are hungry. If you run an informal experiment with your own dog, I’d love to hear about it.
Dog Eating Breakfast image from Shutterstock