The question is on the minds of anyone introducing a dog into a family.  Dogs’ wolf lineage reveals many behaviors that no longer exist or remain dormant under ancient genes Preventative measures you can use to assess a dog's killer instincts would include assessing; the dog’s DNA, its background, the breeder’s values, the breeding bitch’s environmental challenges before, during and after birth, and cognitive testing.

When a News story reports on several dogs mauling a child or killing a small dog we know that the dogs must have had excited enough to have stimulated a dormant instinct to hunt smaller prey.
 
We also hear of dogs biting children, which is likely to have been a warning to the child that went too far.  According to Karen Overall (Dipl. ACVB) and Molly Love (IIABC, AABP), in their paper, “Dog bites to humans—demography, epidemiology, injury, and risk,”  it is clear that boys age 5-9 years are bitten more often than other groups.
 
“In a 3-year annualized study of new dog bite injuries seen in US emergency departments, the highest incidence rate (60.7 bites/10,000 people) was for boys age 5 to 9 years. 3,6,29 Males were bitten significantly more often than were females in all age groups other than human males over 60 years old.6,25 The only exception to this pattern was found on an Indian reservation where dogs were neither owned pets nor stray but took shelter where people worked.55 These statistics strongly indicate that some patterns of interaction (possibly including play) between dogs and humans are gender-biased and that some aspects of these interactions may be conducive to aggression.”
 
From a dog’s point of view of a bite given to a child, derived from our knowledge of wolf to dog evolution, and current scientific studies, they likely were correcting poor behavior with a nip. A child was possibly viewed as being impolite, or disrespectful, or simply not paying attention to social cues sent by the dog. Being impolite to a dog may be reflected in a child who has reached for the dog's toy, or removed food from its bowl,  or perhaps pulled their tail after the dog signaled it was not in a play mindset.  

A bite that punctures skin may still be the dog's attempt at correcting bad behavior in the child, and then further challenged by a dog who has poor mouth control. Puppies learn 'mouthing,' a form of biting not meant to hurt, but to play and correct.  ‘Mouthing’ is a critical lesson established during puppy-hood where experimentation on the strength of a bite is prevalent. Hurting another puppy during play through a too-hard bite can be signaled by a yelp, and is usually met with contrition by the biter.  If repeated again too soon, the defender will return a bite as a correction to immediately stop an action which is painful, scary, or in some other way not wanted.  Mouthing capabilities in puppy-hood are essential for mouth control and bite-control. When a puppy does not developed good 'mouthing' capabilities, their bite may break the skin, without any intention to hurt.
 
When we consider a family being fearful of a dog threatening their children, often breed selection is used  to assess a dog's aggressiveness. Dogs like the Pit Bull or the Doberman would be perceived examples of aggressive dogs. Unfortunately the breed of dog has very little to do with unpopular tendencies which vilify certain breeds.  Dr. Hare's studies on dogs and their cognitive abilities have definitively shown that a dog’s breed does not correlate to aggression.  A dog's breed is not relevant to how one individual dog collects, processes and acts on information like aggression or even submission.  There is significant variation of personalities, behaviors and mindsets within breeds.  Popular ‘kid-friendly’ dogs like Schnoodles can be just as imbalanced and as a Bull-Mastiff, and a Maltese and a Doberman can be cognitively assessed to be highly calm, gentle, submissive and confident.  Breeds provide us only with the original intentions of dogs’ role, hunting and retrieving, and pulling and carrying, but linking behavior to a breed is a myth and should be avoided.  It is the breeders who decide on the generational culling of cognitive traits, such as aggression or friendliness.  Whether a Poodle or a Great Dane, a breeder with a design focus on delivering aggressive dogs, are the real threat to society.

Your family dog with a long caring and gentle lineage, and children under 10 who are monitored and trained in dog's signals is the perfect combination to avoid accidents. If you are adopting a dog, make sure you invest in collecting as much information as possible about the dog.  A visit to the Vet for a full check-up for health conditions, interviews with past owners, an emotional and cognitive test of the dogs to assess personality strengths and weaknesses, will help you to ensure you can minimize risks. 
 

A twitch of the ear, a raised eyebrows, the widening of eyes and the swing of the tail.  A type of canine sign language slowly emerged in my consciousness after a lifetime of being with dogs. It was not a straight-forward path. In some ways it was learned through osmosis, in others through direct observation. There were lots of misunderstandings as the language of dogs is difficult to reciprocate.  Unlike American Sign Language, which creates a dialogue through the exchange of hand gestures, human physiology doesn’t allow for signals to be exchanged as easily with dogs.  Over the years, the language of dogs was further enriched by realizing there was a difference in dogs intelligence found in the way they use inference and logical deductions to problem solve.
 
An early example began fifteen years ago, when we adopted a 6-month old black Labrador, named Raven.  Raven was a black dog, with difficult to distinguish ears and eyebrows due to the uniformity of her color. The only other color distinguishable was her brown eyes.  Her tail, which was more easily noticeable was stiff and not often used. Raven’s face changed as she reached her 4th year, with white hair showing up on her chin, and the most remarkable big bushy white eyebrows. Her ears became outlined in white as well. Eventually her whole face went white and it was during this period, from four to fourteen years, we realized there was much she had been trying to communicate.  The lift of one eyebrow combined with a quick glance at an object, was a request for permission to play with an object. The ears moving back and closer to her head meant she was nervous. Her ears set forward demonstrated a level of eagerness and attention. All of these signals were not received in the early days because it wasn’t noticeable, and we didn’t even know to look for it.
 
Raven first arrived straight from a family who had been bringing her up in a small apartment, too small for her.  Raven was to be a companion to our other adopted 4-month year old dog of unknown origins, named Sheba. On her first night with us, Raven was placed in an adjoining kennel with Sheba. In the morning, the puppy had ripped up the carpet at the bottom of her crate, a sign of what was to come. This unwanted behavior continued for the next 5 weeks. Raven would bite my hand, consumed an entire thick wool mitten, and devoured a leather leash while being worn by Sheba. One day it suddenly stopped.  It wasn’t until several years later, with the distinctive whitening of Raven’s face and recognition of what the signals meant that she had gone through the grief cycle during those first weeks.  
 
Raven was a highly empathic and intelligent dog and I had no idea until her features became more distinctive. 
 
Sheba, on the other hand, provided lots of signals. She was highly vocal, and would stand in direct line of sight, making small guttural noise until she commanded attention and allowed me to resolved her problem.  Sheba would indicate lost toys she could not reach by running back and forth between where she thought the toy was and where I was. This often resulted in the discovery that the toy was not actually where she thought it was and then the requirement to find it before she would settle.  Sheba in comparison communicated in a clear manner.
 
During their early training days, when Raven first joined our family, Sheba appeared delighted to help Raven understand training on the commands she knew. Completing the command she would stand in front of Raven, nose forward, encouraging her to give it a go. Raven learned extraordinarily quickly, where Sheba required many hours of practice before mastering a command. Raven’s intelligence allowed her to eventually far expand Sheba’s capabilities and we were able to put together 3-word commands for specific actions, like “Go, Get, Toy” and “Go, Get, Stick.”  Each word was on its own, a separate command: “Go” meaning move away, “Get” meaning bring me what I am pointing at, and “Toy” or “Stick” was differentiating different objects by name.
 
We went on to develop several training strategies to quickly develop Raven capabilities. An example would be found in one short training lesson where Raven learned to find and bring me my slippers. This required Raven to understand my slipper was not her toy, to recognize, without gesturing, the difference between my slippers and a toy, and to search for a slipper, and deliver one slipper at a time.  While I often received two different slippers, this command brought us many years of joy and she often engaged me by bringing me a slipper without being asked to simply please me.
 
Raven provided me insights into dog’s intelligence and lit a passion for understanding dog emotion. Sheba allowed me to compare the differences between dogs and the way they communicate.  The two dogs provided me with attunement to variations in dogs and their intelligence. I find myself observing dogs in dog parks for body language, vocal messaging and attunement to their owners and other dogs.