There is such a deluge of dog training information online, it's almost impossible to know what to believe! The truth is, animal behavior has been studied as a scientific discipline because it is objectively measurable: you can quantify behavior and leverage statistical analysis. This allows researchers to make reasonable generalizations, and/or specify the likelihood that what they have observed is causing behavior and is a real phenomenon (rather than simply due to chance).
What happens when we apply behavior science and professional dog training know-how to common dog training myths?
"Little dogs are more aggressive because their bad behavior is tolerated or dismissed as cute and harmless by their owners. Little dog owners need to stop 'bubble wrapping' their dogs and correct their bad behavior!"
How many times have you been lovingly greeted with a body check by a bouncing, overly-enthusiastic Labrador or golden retriever? How often have you observed a large dog walking their helplessly protesting owner down the street? Given the prevalence of these and other challenges with "basic training" for large dogs, it may strike you as odd to assume that those same dogs' owners could have somehow acquired superior training skills to eliminate all growling, barking, or biting. People can't even keep their dogs off the sofa; how likely is it that on average, large dog owners possess both the skill, knowledge, and dedication to successfully ward off aggressive behavior simply by virtue of the fact that their dog is large? (By the way, of course small dogs exhibit all the same problems listed above - this post is about being objective, after all!)
Basically, the same owners that suggest they've properly trained their large dog not to exhibit aggression probably can't stop their dog from barking at the door bell. So why are many large dog owners convinced their dogs are generally less-aggressive and better behaved?
Actually, it is pretty clear this is just a misperception or bias, because there is not any data (data being distinct from opinions, anecdotes, or poorly-designed research results, among other things) we are aware of that confirms that all else being equal, smaller-sized dogs are more aggressive than larger-sized dogs. Similarly, the suggestion that small dogs should not be "bubble-wrapped" is at best unfounded, and at worst, dangerous! This is because in the vast majority of cases, the root of aggressive behavior is fear. (If you think about it, you'll realize this is true of any animal, but humans are perhaps the best example.) I often wonder if this may in fact be where the myth comes from in the first place. After all, there is a lot less in this world for a 90 pound dog to fear than a 15 pound pup. For example, certain interactions that a large dog doesn't mind are more likely to be intimidating to a smaller dog. Standing beside a Laborador to pet it is probably more likely to be perceived as a safe interaction, while in contrast it is difficult not to loom over a small dog when trying to pet them, and reach over its head. Therefore, I wouldn't be surprised if more people felt that they could recall experiencing aggressive-looking behavior from smaller dogs than larger dogs. But it's probably not because large dogs are inherently less aggressive, and more likely because small dogs experience the world differently than larger dogs.
So while no one can say for certain that small or large dogs are categorically more or less aggressive, we do know what creates aggression generally - bad experiences (and this may be more likely to occur to smaller dogs). Because all dogs learn the same regardless of size, the reality is that we need to ensure they all are socialized adequately through good experiences - even bubble wrapping them as needed. If we do not, they may learn that the world is a terrifying place, and react fearfully with behaviors intended to increase distance between them and what they fear will cause them harm.
Breed is a reliable predictor of temperament: "that kind of dog, (*insert breed name here*), misbehaves and acts aggressively."
This misconception has been most famously applied throughout the decades to Rottweilers, Doberman, German Shepherds, and "pit bull type dogs", but also relates to Myth 1 above. Actually, despite the dutiful AKC dog show "breed standard" announcer's attempts to wow the crowd with their knowledge of Great Danes' austere character, we do not have any reliable data to suggest that knowing a dog's breed can give you certainty about how they will feel, behaver, or react. (Actually, we have a lot of data to suggest that we don't even know what breed most dogs are! Try it for yourself: https://iaabcprojects.org/quiz/ ) The closest thing we have may be the American Temperament Test Society's statistics ( https://atts.org/ ), which would shock more than a few owners if they consulted the results and saw the maligned breeds listed above scoring quite reliably as well-tempered breeds. While this is the best we have for now and is an admirable attempt at systematically quantifying dog behavior, it is a deeply flawed endeavor from a methodological perspective. For example, sample sizes are too small nearly across the board and do not justify generalization to the larger population of the breed as a whole. They are also hardly unbiased or randomly selected: owners self-select into the testing, and arguably arbitrary exceptions are made for dogs "trained in protection sports".
More importantly, when we condemn "breed" as the culprit, the real culprit sneaks out the back door: human intervention in dog socialization, breeding, and training. Contrary to popular belief, while we don't know too much about the ancient wolves that most dogs descended from an estimated 10,000 years ago, we do know that wolves today rarely, if ever, fight other wolves (it isn't a great evolutionary strategy), and that domestication appears to have created this tendency in modern dogs. We also know that no matter how much you interact with a wolf from birth, it will never prefer humans nor trust them above other wolves. All canids (among myriad other animals) exhibit a "critical period" during which they must be positively, repeatedly exposed to any stimuli they should tolerate or enjoy as adults. For the wolf, this seems to be about 19 days, and intensive human interaction during this period still doesn't change their affiliation potential with humans. By contrast, this period lasts roughly 12 weeks for dogs, and can make all the difference in their behavior as adults. Similarly important is another trait learned early in life: "acquired bite inhibition" (ABI). ABI determines how hard a dog (or any animal with canine teeth) is likely to bite when experiencing intense physiological arousal (fear, rage, etc.), and is learned roughly within dogs' first 16 weeks of life . Barring any genetic issue, the first months of a dog's existence are the top determinants of its temperament.
In sum, we don't have really reliable evidence to allow us to make broad generalizations about breed temperament, and the best data we do have suggests the dog breeds we malign the most may in fact be the best tempered. However, we do know that what really matters is the combination of genetics and early socialization and training in increasing the probability of creating dogs with temperaments humans prefer - regardless of breed.
"Like humans, all dogs are different. Some dogs need a firm hand and dominant owner to learn to behave. Different strokes for different folks!"
While one of the best things about dogs is their unique and varied looks and personalities, this diversity belies the undeniable reality that they're virtually indistinguishable genetically, and that the structures of their brain do not vary materially when it comes to learning! We know that dogs learn primarily two ways: operant and counter conditioning. These are reviewed in more detail in previous posts. Suffice to say here that how best to get behavior is not a matter of opinion, but a matter of fact, and can be subject to scientific inquiry. The same knowledge and processes that allow humans to send astronauts to the moon have made fairly quick work of animal behavior: to train a dog quickly and reliably without eliciting fear or aggression, proper selection of motivators and their application via the appropriate reinforcement schedule using techniques that eliminate pain or force is best. It's actually difficult to understand where the suggestion that a dog needs to be dominated to behave well came from. For example, it's often said that dogs need to walk beside, or ideally behind, their owners. In reality, some of the world's best trained dogs have traditionally worked out front: a hunting dog is expected to quarter in front, an agility dog is sent out ahead to take obstacles, a herding dog pushes livestock before the shepherd, and scent detection dogs guide their handlers to the source. While companion dogs may not have as important of tasks to perform, it seems safe to infer they're not likely to go rogue if their four legs naturally carry them faster than a human's two.