Types of collars

Harnesses and Collars

“Evolution did not necessarily equip dogs to be made captive to collar and lead.”
~Per Jensen
Types of collars
Nylon or Leather Buckle Collars 
This is the most basic type of collar and will help your puppy get used to wearing something around his neck. The name tag and rabies information typically hang from these collars and leashes clip to the D-ring for walks.
These collars are safe provided that the dog does not pull against them. If your dog pulls against them it can cause tracheal damage, vertebral damage, and increased ocularpressure. 
Choke, or Strangle Collars
Choke, or strangle collars (sometimes called chain collars) are metal collars with interlocking links. People who use these collars correctly apply pressure to the base of the dogs neck to check the dogs behaviour. This is a popular method of punishment based training. Used properly, these collars can bring a 96% chance of tracheal, and by extension, esophageal damage.(1) Unfortunately, most people who use these, use them wrong. This makes the chance of injury much higher. Other injuries can include, but are not limited to; vertebrae damage, Spinal cord injuries, hind leg ataxia, transient foreleg paralysis, whiplash, (2) organ damage, blindness, bruising of the larynx, fainting and laryngeal nerve paralysis. (3)
Prong, or Pinch Collars
A prong collar is a metal chain collar with metal points that point inward to the dog’s neck. If the dog pulls against the collar or if you pop the collar, considerable pressure is placed on the dogs neck. Prong collars are popular with punishment based trainers. These collars are preferable to choke collars as the metal points keep the bulk of the pressure off the trachea. However, this product is not without health risks. Though the prong collar only boasts a 16% chance of tracheal damage, the risk is still there. There has been at least one case, a Weimeraner in Germany, where the prong actually punctured the trachea! (4) Additionally, the pressure caused by the prong collar can cause esophageal and spinal damage as well. Unfortunately, due to the route a dog’s optical nerves travel, Prong collars may also cause blindness in a dog.(5) They’ve also proven to cause thin corneas, glaucoma, and corneal lacerations. (6)
Dogs trained on prong collars exhibit greater levels of stress while on walks. Moreover, that stress extends to later walks, even when the prong collar is not present.(7) After just one use, the dog learns that the presence of the owner on walks means pain. Pretty tragic considering we are supposed to be their protectors. There is also increased risk of aggression using prong collars. Some dogs associate the pain they feel with whatever is present in their surroundings. If you popped the collar near a child, for instance, that dog may learn that children cause pain. This could lead to a dog becoming child aggressive. There is a 40% chance of aggression using prong collars. (8)
Also called combo collars, limited or partial slip collars, or Greyhound collars. Martingales are half nylon, half chain. These collars provide less pressure than a choke collar but more than a nylon collar. Martingales are great for dogs with narrow heads as the collar can tighten when needed to keep a dog from slipping free.
Some people use these for training and they use them the same way as one would use a choker. This is not recommended, as when it is used that way it carries the same risks of tracheal and vertebra damage as a choker does.
Head collars
Head collars (Haltis or Gentle Leaders) are collars that go over the dogs muzzle and buckle behind the neck. They are designed to lead dogs by their heads. They are often used for dogs who pull a lot or by people with disabilities.  Many dogs will react negatively to the head collar. Some will buck wildly when it is first put on, others will posture (lower their back and ears). Some dogs get used to it and can enjoy a nice walk with their people, other dogs will continue to buck. If your dog can not get used to the head collar, I strongly recommend discontinuing its use.
There were a couple of studies performed on the head collars. The first took place in 1998 by the Department of Physiology College of Veterinary Medicine at Cornell University. This study compared the head collar to the nylon buckle collar. Physiological and behavioural responses were compared with measurements testing blood pressure, heart rate, respiratory rate and pupillary dilation. Surprisingly, results showed no difference in the physiological response to the types of collars. Behavioural results showed that dogs were more disobedient while wearing nylon collars, but pawed their noses more and watched the handler less while wearing a head collar. Surprisingly, despite posturing and nose pawing, physiological tests show that dogs are not under increased stress while wearing a head collar.
The other study was performed in 2002 by Department of Small Animal Medicine and Surgery, College of Veterinary Medicine, College Station, Texas. This study compared the different brands of head collars and the dogs overall acceptance of the product. This study shows that with time, dogs become accustomed to the head collar and have no preference in one type of head collar over another.
While both studies suggest that head collars are safe for everyday use, if you are uncomfortable with your dogs reaction to this product then I urge you to try a regular collar or a no pull harness instead.
Harnesses go around the chest of the dog, these are safer alternatives to collars and some people feel it gives them greater control. Traditional harnesses encourage pulling by building the chest muscles and providing resistance. Harnesses can make dogs pull more efficiently, however, if a dog is properly trained not to pull then this is rarely a problem. For toy breed dogs and dogs with neck or back problems, harnesses are a much safer alternative to collars.
No pull harnesses
These are types of harnesses that physically, yet gently stop a dog from pulling. These alternatives are safe for both the person and the dog. Some are designed to lead him from the front (as opposed to the back) others put gentle pressure under the dogs forelegs and prevent them from pulling. The idea behind any training tool like these, is to eventually get the dog out of it. They work well for those who need them; they provide no additional stress to the dog and keep the dogs under greater control than traditional collars, prong collars or chokers.
Shock Collars
Also called electronic collars or e-collars (not to be confused with Elizabethan collars, the funnel type collar used to keep a dog from licking his wounds). Shock collars are used to train dogs to hunt, prevent them from pulling, keep them confined, prevent them from barking, and some people have attempted to use them to teach a dog not to pull, jump, or engage in other nuisance behaviours.
Shock collars can cause heightened fear and aggressive response in dogs. There is a 40% chance that a dog trained with a shock collar will become aggressive. Advocates of shock collars claim they do not hurt. This is false, in order to work it has to hurt. It has to hurt badly enough that the dog will think twice before engaging in the behaviour that caused the shock. Unfortunately, dogs do not always know what caused the shock, often it seems random to them. In fact shock collars cause enough pain and controversy that they have been banned in 13 countries! This is because it meets all the criteria for abuse.(9)
There is a well known study done by Richard Polsky in 2000 that showed dogs kept in invisible fences showed a high risk of aggression towards humans. Psychologically, electric fences are no different to a dog than being tied to a tree. Dogs are subject to stimulation, but they can’t reach it. This sets off their frustration drive which is tied to the rage function in the brain. Additionally, since there is no visible fence to protect the dog, the dog has only himself for protection. There is nothing stopping another animal or a person from coming in and harming your dog. There have been cases of dogs killed by birds of prey or other dogs when they were trapped in the fences with nowhere to go.
Other studies (10) have found a rise in stress hormones in dogs who have been trained on shock collars. Dogs associate their owners with pain, this association lasts even when the dog is not wearing the collar. This study found that dogs have higher stress even then dogs who were abused.
Physical risks of shock collars include third degree burns, elevated heart rate and a risk of seizures. 
Bottom line: there are two types of shock collar advocates: those who do not understand them, and those who profit from them. Do you really want to risk your dogs health and well being for somebody who’s in it for the money?
A Special Word on Shock Collars, Prong Collars and Choke Collars.
I previously mentioned the physical dangers of the various collars, but I only touched on the psychological dangers. The reason for this, is because for all three of them, they carry the same psychological risks. In the interest of efficiency, and to make things easier for you to read, I put them together here rather than list them separately in the above columns. I will now expound on the emotional and psychological problems that these collars cause.
The hippocampus of the brain, is the region where anxieties, fears and phobias are stored. Logically, when an animal is introduced to a stressful or painful environment, they will react with fear or withdrawal. Any kind of repeated reinforcement (whether positive or negative) will produce better, more efficient and more numerous connections between neurons. When this kind of stimulus continues, learning will occur in the lateral amygdala. This is one model for learning contextual fear. (11) 
When exposing an animal to an electric shock or the pop of a prong or choke collar, we may also inadvertently be changing other behaviours or processes as well.
There was a well known study (12) that involved training German Shepherd Dogs to be guard dogs. Dogs that were shocked in training but not shocked during later evaluations held their ears lower, and exhibited more stress -related behaviours compared to dogs who were trained without the use of shocks. These differences continued throughout training and on into fieldwork. When the owner or handler (i.e. the person administering the shock) was present the physiological changes in the HPA axis were the most profound.
What we can deduce from this is that A) this type of training is unnecessarily stressful, B) it is painful and C) dogs learn that the presence of his owner and the commands becomes a cue for the shock.
This study is the most widely one used to day and it teaches us that while dogs trained by electric -static shocks can become guard dogs, their behaviours towards humans have changed, they’ve become uncertain and over-reactive. A combination that often proves dangerous to both dogs and humans.
Aroused States and Physical Violence
It is well-known that impulse control functions of the pre-frontal cortex is coordinated with the amygdala-mediated emotional interpretation of social and environmental stimulus. Sudden change that requires rapid assessment of a threat has the potential to create an aggressive response. Aggressive dogs frequently display exaggerated autonomic responses to threats including excitability, panic and fear.(13)
There is a well documented link between dog abuse, spousal abuse and child abuse. Studies that investigated domestic violence showed that the perpetrators experience intense autonomic arousal and symptoms of panic at the moment of the attacks. They experience heart palpitations, increased respiration, tremors, fear, and feelings of losing control. Dogs who are in states of arousal also exhibit signs of increased blood pressure and heart rate. The similarities between the two can not be overstated.
Police officers who have responded to calls of domestic violence know not to use force or pain to diffuse the situation. In such states of arousal, the perpetrator will become more unpredictable and dangerous.
Dogs, whose brains and bodies are under the same signs of stress and arousal as humans while aggressing, can not be handled. Hopefully by now, everyone knows not to break up a dog fight using your bare hands, in that heightened aggressive state, the dog will not know it is you and bite back. Using a shock collar or prong collar to “control” an aggressive dog causes the same problems as trying to physically interfere with a fighting dog. In that heightened state of arousal the dog becomes more angry.(14) One simply can’t use force and violence to stop violence. Our brains aren’t set up to work that way.
There are better and safer alternatives to aversive devices, that have been proven in the long run to work even better. There is no need to rely on outdated methodology when science has provided us with more effective and safer means of training. Canine behaviour, genetics and chemistry are growing fields of ethology and science. It is better to be on the cutting edge of science then left behind. 
Whichever type of collar you decide to use, collars should only be used to attach the leash and the dog tags. Collars should never be used for training. Not only does using a collar for training increase the chance of serious health risks, but it teaches the dog that the collar is in charge, not you. If the dog only responds if he is wearing a collar then that means that he is not trained to you. A dog should respond the same regardless of whether or not anything is on his neck. For dogs that pull, try walking them off lead in a fenced in area. This teaches them to pay more attention to you on walks, no matter what they may be wearing.
There is a lot of controversy surrounding collars and harnesses. Most people want to do the right thing for their dogs and it can be confusing knowing exactly what is right for your particular dog. There are many people out their with very good intentions who will advise you what you should use to help control your dog. If your friend, coworker, trainer or family member recommends something, thank them for their advice, but if you do not feel comfortable doing something, don’t do it. If the person advising you can not back up (with peer reviewed evidence) what they are trying to sell, then it may not be something that is right for your dog.
At the end of the day, he is your dog. He is in your care and you must do absolutely everything you can to care for him properly, for he trusts you so completely.
I do not have the luxury of basing anything on opinions, too many people count on me for the safety and well-being of their dogs. Everything written above is falsifiable (can be proven) and based on peer-reviewed studies done by researchers. I have provided the references below, I urge you to study for yourselves, and if you have any questions, please don’t hesitate to ask.
In 2012, an experiment was conducted in Switzerland that tested shock collars on people. You can find an outline of it here. 
They link to some videos on YouTube of their experiments. I find it interesting that the same visible signs of stress on the test subjects are quite similar to the visible signs of stress I’ve witnessed on dogs during their training sessions.
The test is interesting both as an insight in to shock collar training for dogs but also as possible insight into human behaviour as well. If there are any psychologists reading this, I’d love to hear your two cents. 
dog training history
1.Neurologist Jean Zuniga, MD. PhD
2. Bremmier et. al. 2000
3,4. Per Jensen
5. Dreyton Michaels PhD.
6. Pauli, et al. 2006
7.RVincent and AR Mitchall  
8. RL Johnson (1976)
9. Dr. Karen Overall, PhD
10,11. Shalke et.al.
12. Shilder and van der Borg 2004
13. Steven R. Lindsey
14.Dr. Karen Overall Ph.D