Now, that being said, while breed doesn't tend to communicate that much it seems. At least based on what Citizens Scientist are finding as they play Dognition. It's not that with the sample size that we have, we can't see really interesting differences between different breed groups. And here's one of the findings that we made that we wouldn't have been able to see otherwise, because you need a lot of power to be able to see differences like this. In all the ten games that you can potentially play with your dog through Dognition, only two did we find significant differences between different breed groups. And K is where we saw the big differences, where the arm pointing game and the memory versus smell game. And the differences, I think, are so interesting, because it really highlights what we talked about. And I think it really illustrates beautifully what a cognitive approach is, versus our traditional way of thinking about animals either having more or less intelligence. What we found was that purebreds, dogs that people registered in the Dognition database as a purebred dog, they tended to be much more reliant on human gestures than mixed breed dogs. So, you might conclude then that mixed breed dogs are less intelligent. Well except that when you look at a different skill, mixed breed dogs are actually much more reliant on their memory, than pure bred dogs are. And so, who's smarter, the dog that relies on their memory more, or the dog that relies on human gestures more. I think this just beautifully illustrates that there probably are breed differences. They're a bit subtle in some cases, and we're gonna need a very powerful data set. That means lots and lots of dogs to be able to identify them. And hopefully as more citizen scientists contribute, we'll be able to have more resolution. And not just look at breed groups, but even look at specific breeds. And compare all sorts of things that everybody's really interested in. And hopefully there will be lots of application of that as well. So I think if you're not convinced already that there might not be huge breed differences, or at least they're not so robust that they're easy to see, I can tell you about work that we're doing at the Duke Canine Cognition Center, comparing Labrador retrievers. We have Labrador retrievers that work as assistance dogs trained by K9 companions for independence. And we also are working with the Office of Naval Research and we're working with Labrador retrievers that are trained to detect explosives, or improvised explosive devices in Iraq and Afghanistan. And what we found is, even though they are all Labrador retrievers, every single dog that we have studied, and we have data on hundreds of them. There are huge differences, cognitively, between Labrador retrievers that are being bred and trained to be assistance dogs, and Labrador retrievers, again, the same breed, that are being bred and trained to be dogs that detect explosives. So don't worry about this figure, but what you see in the words on the left side are actually different games, or cognitive games that we've played with these dogs. And every game that has a bar above that dotted line means that that's where the two populations are different. And if the differences are significant. And what is really interesting about this data is that, of course, it's within the same breed. And what it means is that there is so much variability within a breed, in terms of the cognitive ability, that breed itself is not communicating anything reliable about a dog's cognitive ability. If you can have that kind of difference within a breed, then it's unclear what breed really can tell you about your dog's cognition. The only way to know your dog's cognitive profile, essentially, is to measure it. Okay, so to sum up what we've learned about breed differences and their cognitive abilities. At this point, is that breed doesn't seem to communicate much about cognitive ability. I do think it tells, I'm sure it says a lot about temperament. I'm not saying that breeds aren't different. I'm sure that they solve problems differently, but I imagine a lot of that difference is going to be relevant more to their temperament and probably their motivation in many cases. Dognition profiles have a much more power description of a dog's cognitive style than using breed. Citizen science data from dognition has revealed that pure breeds are more reliant on human gestures than mixed breeds. But that mixed breeds are more reliant on their memory than pure breeds. And there's tremendous individual variability within a dog breed, in terms of temperament behavior but especially cognition. So that makes talking about a breed specific cognitive profile very difficult. Especially when you think about the example of our own research where we've compared two populations of Labrador retrievers that are so different from one another, even though they are absolutely the same breed.
