Are mixed breed dogs like the Cockerpoo and labradoodle REALLY more badly behaved that pedigree dogs?

You might have seen the headlines over the last few days suggesting that so-called “designer crossbreeds” are more poorly behaved than their purebred counterparts. It’s one of those claims that moves quickly because it feels simple, it feels like it offers an explanation, and it gives people something quite easy to latch onto. But the more I’ve looked at it, the more I’ve found myself coming back to a much simpler question, and it’s one that doesn’t seem to be asked nearly often enough.

What exactly was measured?

Because when you strip it back, what sits underneath this isn’t observed behaviour in any real sense, it isn’t controlled assessment, and it isn’t independently verified in a way that most people would probably assume when they read something described as research. What it actually is, is an owner-completed questionnaire. And I think it’s important to say at this point that I’ve taken the time to go through that questionnaire properly, to read the questions, to look at how they’re structured and what’s being asked, because that’s not something most people will ever do. Most people will read the summary, or the headline, or the conclusion, and take it at face value.

And that isn’t to say there’s no value in owner surveys at all, because there is. If you’re asking people what they feed their dog, how often they go to the vet, how frequently the dog is exercised, those are relatively straightforward bits of information. You’re essentially just collecting data, counting, building a picture of how people interact with their dogs in a practical sense. There’s nothing particularly complicated about that.

But the moment you move into behaviour, you’re in a completely different place.

Because now you’re not asking people to report something factual, you’re asking them to interpret. You’re asking them to decide, in their own mind, what counts as good or bad behaviour, what’s acceptable, what isn’t, what’s normal, what’s a problem. And you’re doing that without any shared standard, without any real baseline, and without any way of checking whether what they’re saying actually lines up with what’s happening in front of them.

So what you end up with isn’t a dataset of behaviour in any objective sense, it’s a dataset of interpretations. And those interpretations are coming from people with very different levels of experience, different expectations, different tolerances, and in many cases, quite a limited understanding of behaviour itself. Most of us as dog owners don’t actually understand behaviour as a subject, and that’s not meant as a criticism, it’s just the reality of it. Even within the industry, plenty of dog trainers and vets struggle to fully understand the motivations behind behaviour and to interpret what they’re looking at accurately.

And when you look at how these surveys are actually distributed, that becomes even more relevant. They’re shared through veterinary practices, training clubs, magazines, social media, online links, and participation is voluntary. So at the most basic level, you’re relying on trust. You’re trusting that the person filling it in has a dog. You’re trusting that they’re describing that dog accurately. You’re trusting that they understand the questions in the way they’re intended, and that their answers reflect what’s actually happening rather than what they think should be happening, or what they feel is the right answer to give.

Most people will be answering honestly, of course they will, but honesty and accuracy aren’t always the same thing. And there’s no mechanism within that process to verify or corroborate what’s being said. So the entire thing rests on a kind of collective trust that the information going in is both accurate and comparable, when there’s no real way of confirming that.

That becomes even more obvious when you start looking at the questions themselves. Something as simple as barking at strangers sounds straightforward enough, but in practice it could mean almost anything. One bark, repeated barking, excitement, frustration, fear, something escalating, something that stops as soon as it starts. All of those are completely different behaviours in terms of what they mean and what you’d do about them, but within the structure of a questionnaire they’re all pulled together and reduced down to how often they happen. The context disappears, the reason disappears, what happens next disappears, and once that’s gone, you’re left with something that looks neat on paper but doesn’t really tell you what you need to know.

The same thing runs right through something like recall. You’re asking whether a dog comes when called, but what does that actually look like? From how far? Away from what? With what level of distraction? Is the dog already looking at the handler, or is it fully committed to something else? A dog coming back from a few feet away in a quiet environment isn’t doing the same thing as a dog turning away from wildlife at distance and coming straight back under control, but both of those can end up being recorded in exactly the same way. The word recall is fixed, but the behaviour behind it varies massively.

And when you look at the questionnaire in full, that imbalance becomes quite hard to ignore. Out of roughly a hundred questions, only a small handful—around four—relate in any meaningful way to recall or chasing behaviour. Four questions touching on what, outside of aggression, represents one of the biggest and most common challenges dog owners face in the real world. That alone tells you something about what is being prioritised and what isn’t.

Because for a lot of people, that’s exactly where things fall apart. Not in the living room, not in the garden, but out in the real world, at distance, under distraction, when it actually matters. And yet within the structure of the questionnaire, those behaviours are barely explored.

And then you get to what, for me, is one of the more telling parts of the whole thing, which is how that same area of behaviour is reduced even further. Predatory behaviour, which sits right alongside recall in practical terms, particularly here where dogs are exposed to livestock, wildlife, open land, gets reduced down to a handful of hypothetical questions about whether a dog would chase small animals given the opportunity, and then placed under a section labelled miscellaneous.

There’s no real sense of how far the dog will go, how long it will stay engaged, whether it can be stopped, whether it comes back, what happens once it commits to that behaviour. The difference between a quick chase and a full pursuit is enormous in real terms, but within the structure of the questionnaire, it barely exists.

And when you read through it, you also start to notice that it’s very clearly built around a particular kind of environment. You see references to dogs on a leash, behaviour in the yard, interactions with small animals that reflect a more suburban, often American setting. That doesn’t translate particularly well into other environments, especially here, where dogs are far more likely to be off lead in open spaces, encountering livestock, wildlife, and a completely different set of challenges. So while the tool is standardised, and that makes it convenient from a research point of view, it isn’t necessarily representative of the environments it’s being applied to.

At the same time, you can’t really ignore how the explanation around behaviour seems to shift depending on what’s being talked about at the time. Not that long ago, the increase in problem behaviour was being put down to lockdown dogs, lack of socialisation, limited exposure, reduced access to training. That was the explanation. Now, without much pause, the focus seems to move toward breed type, as though behaviour has suddenly become something that sits inside the dog itself, rather than something that develops through experience and interaction with the world around it.

And alongside that, you see something else happening as well, which is how quickly those conclusions get picked up and repeated. Once a piece of research is released, it’s not that each media outlet goes away, looks at the methodology, reads through the questions, and comes to its own independent view. In most cases, what’s being circulated is a press release, and that press release is repeated, shared, and repackaged across multiple platforms.

So what you end up with is the same message appearing in slightly different forms, across social media, print media, and news outlets, all at once. And because it’s coming from multiple directions, it starts to feel established, as though it’s been independently examined and broadly agreed upon, when in reality it’s often the same source being echoed again and again.

And when that happens, it creates the impression of consensus without necessarily having gone through the process that would justify it.

And I think that’s probably why I’ve found myself writing about it at all.

Because once you’ve actually looked at how the data is gathered, once you’ve read the questions and seen what sits underneath the headline, it becomes quite difficult to read those headlines in the same way again.

It’s one of those things that, once you notice it, you don’t really unnotice.

And there’s more to it than this, but this felt like a good place to stop and at least put that much of what is really happening into words.

Jamie Penrith
20th March 2026