In the 1970s, a psychologist tested the self-restraint of preschool children. Each child was offered a marshmallow. The children were told that they would get two marshmallows if they could delay eating the treat, and then left alone in a room for fifteen minutes. The researcher recorded what happened. The efforts of some children to stare down the treat or to distract themselves from it are both comical and painful to watch. Of course, some children inhaled the treat as soon as the researcher left the room. A recent book (The Marshmallow Test) describes this experiment and the follow-up studies of those children. The marshmallow test and other research on the ability to delay gratification shows that those who can exercise self-control in the face of temptation have better “life outcomes,” as measured by a variety of criteria, including SAT scores, social and cognitive functioning, long-term health, and retirement planning.
What does all of this have to do with thinking dogs?
Alberta experienced her own version of the marshmallow test recently. To say that Alberta loves treats is a bit like saying that I love chocolate. Alberta not only loves treats, she is not terribly fussy about which treats she gets. For sure, there are better treats, for example this bison and beef jerky concoction that I get at Costco and that, for some reason, Jana, Cali, and Alberta will do anything for. But ordinary, boring biscuits are fine too, and they are happily accepted as rewards for a job well done.
In her guide dog work, Alberta comes across many items that fit this dog’s definition of “treat,” and she works very hard to resist bits of food that just happen to be lying on the floor.
Alberta is justifiably proud of her hard-earned restraint, but more importantly, she wants Deni to know. So, in the course of a day’s work, if Alberta sees food on the floor and gives it a wide berth, she also nudges Deni to make sure that Deni knows just how good she is being. She pushes Deni hard with her nose, hoping that Deni will notice the ignored object. She often nudges Deni right near the pocket where Deni keeps the dog treats, just in case Deni might want to reward this extraordinary show of restraint. A girl can hope, can’t she?
Alberta knows the rule that she can’t grab food off the ground when she’s working. She wants to believe that that rule does not apply when she’s off-duty (her harness is off). She also knows that, even while working, she’s allowed to take treats that Deni hands her for particularly notable service. But she recently encountered a situation that blurred these lines a bit, a marshmallow test for dogs. Her reaction was remarkable.
Guiding Deni down a street in Saugatuck, Michigan, Alberta (along with her entourage of two other human family members) passed by a store that not only had a full doggy water bowl sitting by the sidewalk, but also a full bowl of doggy cookies. Just sitting there for the taking. An open invitation. Irresistible.
Alberta headed for the water, took a drink, noticed the cookie bowl and … stopped dead. Confused. She looked at the biscuits. Looked at Deni. Furrowed her brow. Looked longingly at the treats. But she did not touch the treats.
We’d all stopped to watch the unfolding marshmallow-like drama. Alberta really wanted to gobble up as many of those dog cookies as she could. But she did not take one. She did, however, look at every one of us to make sure that we all knew how good she was being. Deni rewarded her by picking up a few biscuits from the bowl and handing them to Alberta.
Watching the interaction, I got to thinking. Most dogs who walk by this shop are not trained service dogs. Though many, like Jana and Cali, have had some training and certainly know that they are not supposed to just devour everything in sight, they don’t always have the restraint to follow through. Having more than once found myself alone with a new box of Trader Joe’s dark chocolate peanut butter cups, I can relate.
I wondered how often a doggy passerby just digs in and eats all the cookies in the bowl. How many battles between hungry hounds and their hapless handlers has the shop owner witnessed? Does the handler ever win? And really, what was that shop owner thinking?
But back to Alberta. In need of photos for this blog post, I asked Deni to re-create Alberta’s marshmallow test. The photo gallery (presented in order) at the top of this blog post shows that, like the successful children in the original marshmallow test, Alberta devised a series of ways to distract herself. Some of the children looked away, as Alberta did. Some sang songs or recited the alphabet. Alberta did neither of these. Some children closed their eyes. Upon realizing that, even after she had turned away, the bowl of biscuits was still there, Alberta closed her eyes.
Alberta has not only learned to resist random bits of food on cafeteria floors, sidewalks, and the like, but there she was, on that Michigan sidewalk and again in Deni’s office, passing up food that was obviously meant for dogs, placed there for her enjoyment. This shows us that dogs are able to do some high-level thinking and processing.
If Alberta were purely instinct-driven, that bowl would have been emptied in seconds flat. If she were operating only out of fear of punishment or hope of reward, she might have surreptitiously sneaked a mouthful of biscuits before Deni noticed, just to see if she could get away with it — and been rewarded by the treat, even if she got scolded after. But she went beyond a gut-instinct response and even beyond the basic (low) level of moral development that governs much of human and animal behavior. She paused, checked in with Deni, and did the right thing — even though she really wanted those cookies. We’re eagerly looking forward to seeing Alberta’s SAT scores and are consulting her for retirement advice. But in my next post, I will describe more practical ways we can apply the doggy marshmallow test to our relationships with our dogs.
7 thoughts on “Alberta’s Marshmallow Test”
[…] and I were delighted that both Alberta and Koala did well on our cleverly designed doggy version of the marshmallow test — a […]
[…] the equivalent of the Marshmallow Test. I’ve played around with that a bit with Koala and Alberta, though I lack access to an MRI machine. So I was very interested in his findings. It turns out […]
[…] devised a dog version of the marshmallow test for her guide dog. She administered it to her guide Alberta a few years ago. Alberta did well; she actually did many of the things that children who take the […]
[…] written about both Alberta’s and Koala’s adventures with the marshmallow test and I’ve been thinking about what it […]
[…] Deni’s previous guide dog, as you might recall, did very well on the marshmallow test. She clearly worked to distract herself from the treats, employing some of the methods the children […]
[…] Alberta’s Marshmallow Test, I described Alberta’s admirable ability to resist temptation in the form of a bowl of dog […]
My service-dogs-in-training get this feedback all the time. When we’re just learning to walk on a loose leash, I’m a reward-giving machine. Every couple of steps, a bright “yes” and a couple of kibble as a reward. But the dogs can’t always walk and chew kibble at the same time, so they get dropped. And so we also practice “leave it” when a kibble doesn’t quite get into the dog right. A well-ignored kibble on the ground gets a couple more as a reward.
As far as I know, none have found the strategy of dropping the first kibble to up the kibble-count. But they certainly look expectant when they pass the dropped kibble on the return trip.