DunbarAcademy.com

IS DOG TRAINING BROKEN?

The past decade has seen a massive resurgence in 'quality-of-life-destroying' problems, such as: 1. Fear and Aggression towards People; 2. Dog-Dog Reactivity; 3. Separation Anxiety; and 4. Hyperactivity, Inattentiveness, and Lack of Control when excited (arriving guests) or stressed (unfamiliar dogs). Veterinarians are expressing increasing concern about difficult-to-handle dogs. Basically, we are back to the same worrying problems that we had in the 70s.

The re-emergence of these common, predictable, and preventable behavior, temperament and training problems most likely reflects recent changes in dog training, specifically that we are neglecting the techniques that largely eliminated these problems from 1982 through 2015.

If you are intrigued by any of this, please check out my upcoming seminar in the UK (CLICK HERE) and the US (CLICK HERE). 

1. PREVENTION: Preventative Puppy Socialization with numerous unfamiliar people, Handling, and Environmental Enrichment; Short-term and Long-term Confinement Routines to teach puppies to settle down quietly for regular short periods with food-stuffed chewtoys to prevent separation anxiety when left at home alone for much longer periods, and to facilitate teaching them to eliminate on cue; and Off-leash Puppy Classes to immediately resolve any signs of fear towards people, for puppies to learn bite inhibition and social savvy and so prevent dog-dog reactivity during adolescence, and for owners to learn off-leash control and so, avoid the difficult transition from on-leash to OFF-leash.

2. LURE-REWARD TRAINING for TEACHING off-leash cued-behaviors and TESTING their Response-Reliability (comprehension and motivation) to use as clear instructions prior to task, and for verbal guidance, i.e., instructive, non-aversive 'punishment' for when dogs misbehave or are non-compliant. I think that many trainers have forgotten these techniques, and many younger trainers may not have heard about them.

When I first introduced off-leash puppy classes and positive reinforcement to teach cued responses (lure-reward training) to the world of 'on-leash and leash-correction' dog training in 1982, the techniques were all so old, they were new again. I spent two weeks in the AKC Library reading books written in the 1700s and 1800s, all the techniques were there. Early socialization and handling of dogs, horses and livestock, classical conditioning, training off-leash, luring, shaping, positive reinforcement, verbal instructions, and verbal corrections were the natural way, even though camouflaged by many extremely aversive techniques of the day. It would seem that so many reward-based techniques, and there are many, are again so old that they must become new again. Indeed, we urgently need to re-introduce and re-popularize these approaches for the mental wellbeing of dogs and their people, and to re-invent dog training once more.

1. PREVENTION
The 'Open Paw' Minimal Mental Health Requirements for Raising Puppies
Handling and Socialization with 100 PEOPLE prior to 8-weeks of age safely in the kennel, and another 100 people during their first month safely at home. Outdoor shoes must remain outside.
Environmental Enrichment — expose you puppy to sound tapes, and all sorts of weird objects that cover the floor, such as different substrates, crinkly paper, plastic pools filled with empty water bottles, cardboard boxes, wobbly seesaws, vacuum cleaners, etc., etc., etc. Also, carry, cart, or drive to Main Street or shopping center parking lots, for your pup to see and meet even more unfamiliar people and habituate to the urban hubbub of noises, smells, sights, and sounds of milling people, other dogs, passing cars, motorcycles, trucks, skateboarders. etc.

An Hourly Confinement Schedule during the first few weeks to prepare puppies for lengthier periods of being left at home alone and so, prevent separation anxiety. Moreover, helping puppies settle down promptly, calmly, happily, and quietly for short periods of personal chewtoy-time in a crate or puppy-proofed room teaches the pup to develop a food-stuffed chewtoy-habit which prevents destructive chewing, and makes it quick and easy to teach puppies and adult dogs to eliminate on cue. Feeding mainly from stuffed-chewtoys reduces activity and boredom/recreational barking by 90% within just a couple of days.

You can learn more about the Open Paw program in the Free Course Collection at DunbarAcademy.com, where you may download three books for free, explaining how to create a learning schedule for your puppy or newly adopted adult dog.

OFF-Leash Puppy Socialization & Training Classes

Off-leash puppy classes allow pups to socialize with and be handled and trained by another couple of dozen unfamiliar people (the other pups' owners) and allow trainers to identify incipient signs of fear or aggression and nip them in the bud. (Check the short videos in recent posts on the Dunbar Academy FaceBook page to see how quickly fear is resolved and how to deal with growly and scrappy puppies.) Off-leash classes are essential for puppies to bump-start their all-important play behavior so they may develop bite inhibition (a soft-mouth when play-fighting and play-biting) and dog-dog savvy to give them the confidence and know-how to deal with other dogs in the real world that may not be similarly socialized and friendly and thus, prevent adolescent-onset dog-dog reactivity. Additionally, owners learn to master off-leash control in a very distracting but safe setting. In their very first session, owners learn how to teach their dog to Sit, Settle and pay attention while the other dogs are playing, and how to teach Shush as required.

2. LURE-REWARD TRAINING
My latest book, Barking Up the Right Tree, analyses the pros and cons of the five most common positive reinforcement techniques in terms of ease, speed, and effectiveness for each training scenario. For teaching basic manners, Come, Sit, Down, Stand, Stay, Heel, Hustle, Steady, Find, Fetch, Go to... (people or places), etc., for companion dogs, competition and working dogs, lure-reward training rules supreme.

Lure-Reward Training:
Teaches dogs ESL from the outset — English as a Second Language

• Teaches puppies off-leash from the outset, thus making training a one-step process — instead of training on-leash and then off-leash. (When teaching 'on-leash walking', we first teach dogs to follow off-leash, heel off-leash and then, attach the leash to our trained dog to teach heeling on-leash, prior to ever attempting to walk the dog along the sidewalk on leash. (OOC leash-pullers are an incendiary trigger for other dogs and owners.)
• Systematically and quickly phases out reliance on ALL training tools — food lures and food rewards, and prevents leash-dependency
• Uses a wide variety of considerably more powerful rewards
• Uses much more effective reinforcement schedules
• Creates the highest levels of response-reliability, which are tested before, during, and after each session

Until eventually, dogs become self-motivated and internally reinforced.

Once we have TESTED the dog's level of comprehension of our verbal instructions and their motivation to respond, (Response-Reliability Percentages), we may use a single word as guidance when dogs misbehave or are non-compliant, which makes any aversive means clunky and well... obsolete. For examples, "Sit", "Shush", "Find Chewtoy", "Go to your Bed", "Settle".

Additionally, in Barking Up the Right Tree, I objectively explain, why the use of aversive 'punishment' is very rarely easier, quicker, or more effective than reward-based techniques. The terms 'aversive' and 'punishment' are NOT synonymous and outside of a laboratory setting wherein computers (consistent 24/7, and with impeccable timing) train captive rats (that cannot escape shock), when humans administer aversive 'punishment', very few aversive stimuli actually act as punishment to reduce and eliminate targeted behavior, as evidenced by their frequent and continued use. It requires a very high skill-set for aversive punishment to be administered effectively.

If aversive stimuli were punishments, the targeted behavior would rapidly decrease in frequency until eliminated, which of course, also eliminates the need for further punishment. But is this what we see? When we watch people giving leash 'correction' after leash 'correction' when walking their dog down the sidewalk? Therefore, so many aversive stimuli cannot be defined as punishment. Remember, in the behavioral sciences, punishments, rewards, and trainers are all defined by their EFFECT on behavior (increasing, decreasing, and changing behaviors), and not by their NATURE (pleasant or unpleasant). A punishment can be pleasant. In fact, we have a slew of highly effective non-aversive punishments — verbal 'instructions' and 'verbal guidance' and it would behoove all trainers to use them.

Positive Reinforcement Training is Unwell
Many positive reinforcement trainers dislike the very thought of punishment because they think it must be unpleasant and stressful. Consequently, they are often at a loss for what to do when dogs misbehave or are non-compliant. Turn your back on the dog, (do nothing)? Distract? Teach a DRO or DRI? Again, punishment need NOT be unpleasant, and if you can train a DRI, why not just ask yourself the question, "If this is 'wrong'; What is right?" And then instruct your dog to respond exactly as you would like in that scenario: "Sit", "Shush" "Find your Chewtoy" It's as easy as that; lure-reward training in a nutshell.
But today's dog training ills are much deeper that people's aversion to aversive punishment. Indeed, I think that it is positive reinforcement not being used to maximal effect that is failing dogs and preventing them from leaning at the speed of light. The reason is obvious, rather than using the 'effective practical training of dogs' as a guide, so many trainers are wedded to a Century-old, credal theory of computers training rats in cages, which of course, has dubious relevance to people training dogs off-leash at home, or in parks. When training, we must frequently ask the 'Dr. Phil questions': "Did it work?" and "How well did it work?" QUANTIFICATION is essential for dog training to evolve progressively as a science.

Reinforcement Woes
So many training tools become permanent management tools, such as food, leashes, 'training collars', halters, harnesses, etc., Food rewards especially are overly used in such quantity to drastically reduce their reinforcing power, and reinforcement schedules have never evolved from the constraints of mundane, laboratory CC, FI, FR, VI and VR, which are all fraught with problems — devaluing the reward's reinforcing power, scalloping of behavior, going 'on strike'. Certainly, we have all been taught that a variable schedule is much more powerful than a fixed (predictable and expected) schedule. Remember the textbook comparison between a slot machine and a food-vending machine. An essential fact to learn from science, BUT... from an applied viewpoint, who could administer a variable schedule and train a dog at the same time? Who could even calculate a variable schedule? Go on have a go! Create a VI-5 to reinforce a stay on average every five seconds but with varied intervals between the rewards. Write down ten intervals that average 5 seconds.

But why bother with this at all? Random Reinforcement is just as effective as our vaulted variable schedules, just as long as the random schedule is not too rich. (Always count out the number of food rewards beforehand)! One of the sheer beauties of positive reinforcement techniques: you can be inconsistent when giving feedback. In fact, you should be inconsistent.

Moreover, although lickety-split timing is important when reinforcing short-duration behaviors, such as Sit or High Five, it is not essential. In fact, in the lure-reward training world, we encourage owners to delay giving rewards when shaping progressively increasing durations of Sit, Down, Stand, Bang! Watch, and Off, so that for example, your praise and reward reinforces a quick Sit and a Sit-Stay of a variable, but progressively increasing number of seconds. When reinforcing a multi-minute Stay or Watch, praise and reward whenever you like.

But What about Quality?
All the predetermined schedules (mentioned above), and random reinforcement, reinforce just as many below-median quality responses as above-median quality responses, which is not altogether smart. So, when rewarding at random (zero brain-strain for people), obviously let your naturally-born inconsistencies be influenced by the quantity and quality of responses.
Humans could never approximate the consistency and impeccable timing of computers and so we fail miserably when trying to administer aversive punishment effectively. But no matter.

1. In many ways we are smarter than computers. (Didn't we invent computers? Doesn't AI require 'deep learning'?) When it comes to reinforcement, we are unmatched, because of our inimitable human ability to reinforce the ineffable — style, panache, pizzazz, the Wow!!! factor, the 'je ne sais quoi' — the indescribable attributes of a performance that touch our hearts and minds that come from our lifetime of deep learning.
Rewarding at random quantitatively improves rates of responding just as well as any computer generated and administered variable reinforcement schedule, and influencing our random reinforcement by preferentially rewarding the dog for what we consider to be stellar responses, qualitatively improves responses in leaps and bounds.

2. And when we lure-reward train cued-responses, and routinely test their reliability, speed, duration, and quality, we have verbal instructions that work, which we may use as verbal guidance to eliminate misbehaviors one by one, and resolve non-compliance by using words, and by not even having to raise our voice. At which point ... the mere notion of aversive procedures becomes ... simply unnecessary.  

THE THREE STAGES OF LURE-REWARD TRAINING
Lure-Reward Training is a three-step, comprehensive, reward-training methodology for teaching reliable cued responses, comprising: Systematically phasing out food in the hand (Stage 1) and food in the bait bag (Stage 2), so that compliance does not become contingent on the trainer having food but instead, using the vast array of life rewards that are all around us. In Stage 3, we use our dog's newly acquired vocabulary to: offer specific, often single-word instructions when dogs misbehave, such as Toilet, Chewtoy, Sit, and Shush; and non-aversive, verbal, negative reinforcement when dogs are non-compliant. We have dialogue, so dogs understand what we would like them to do, and differential PRAISE to reinforce the many varied degrees of speed, precision, and quality of responses. Humans have a highly sophisticated language; we should use it to communicate with our dogs.

Moreover, dogs are trained off-leash from the outset, so that we routinely test for the dog's consnet to play The Training Game with us, and the trainer does not have to make the agonizing and tricky transition from on-leash to off-leash.
Lure-Reward Training is quick, easy, and effective, and as enjoyable as all the other reward- based training techniques, in fact, perhaps more so, because now we can talk to our dogs and they understand so much of what we say.

STAGE 1: To TEACH verbally-cued behaviors from the outset and TEST the dog's comprehension of Handsignals and Verbal Instructions, i.e., to PHASE OUT FOOD LURES completely.

Use a food or toy lure to teach the meaning of handsignals via the Basic Training Sequence that comprises the essential components of associative learning and positive reinforcement:

1. Request —> 2. Lure —> 3. Response —> 4. PRAISE and maybe Reward
Then go cold-turkey on food in the hand and use Handsignals (as hand-lures) to teach dogs the meaning of the WORDS we use for INSTRUCTIONS, VERBAL GUIDANCE, and PRAISE.

1. Request —> 2. Handsignal (no food) —> 3. Response —> 4. PRAISE and maybe Reward
FMI: To help you grasp luring skills, see the sequential photographs in my Good Little Dog Book that is available for free download from https://dunbar.info/freepupbooks and the many video courses at DunbarAcademy.com — a video wonderland of dog behavior and training.

TEST RESPONSE-RELIABILITY PERCENTAGES
Basically, RR%s are a single index of the dog's Comprehension of our Instructions and Motivation to Respond. Don't Worry, a child of four could calculate these (with a calculator). Walk round your kitchen and instruct your dog to Sit 10 times using Verbal Cues only. Video, so that afterwards you can count the total number of verbal cues.

The RR% is the Total Number of Responses (10) divided by Total Number of Verbal Cues (e.g., 13), x100 (to convert the decimal to a percentage) = 77%

Easy-peasy right? OK, now repeat the above exercise in the living room, in your yard, and on-leash on the sidewalk and have other family members do likewise. You will find that the RR%s differ considerably according to who was training the dog, where, what was going on, and what the dog was doing each time they were instructed to Sit. Now you know the situations wherein your dog does well, and which scenarios require practice.

When testing, I usually use a simple test sequence — Sit-Down-Sit-Stand-Down-Stand that I repeat three times in each scenario but especially, before and after each training session, and often several times within each session.

For safety, it is important for you to have a realistic view of the precise level of verbal control over your dog in a variety of scenarios before performance in the obedience ring, or in real life, for example, when your dog runs off-leash from car to house, or vice versa, by far the most common situation for dogs to be run over by a passing car.

THE IMPORTANCE OF QUANTIFICATION
If behaviors are not cued from the outset, it is impossible to quantify response-reliability. And
without ongoing quantification of Response-Reliability, Response:Reward Ratios, speed of Sits and Recalls, and duration of stays and eye-contact, trainers cannot assess the speed and effectiveness of their methods and so, they have no goal for improvement — to continually quest to surpass their person bests.

Data ROCKS! A single training session generates oodles and oodles of irrefutable data. Testing just six dogs, is a reputable research study in its own right with results that do not require statistics because they pass the Intra-Ocular Trauma Test — the difference between graphed before and after session scores are so obvious that they 'hit you between the eyes'. These are the very studies that the APDT Foundation should be rewarding — with prizes for completed research on dog training by dog trainers, and publishing the studies in an online, APDT Journal of Dog Training.

If we want dog training to become truly science-based, we must stop quibbling about theoretical and instead, concentrate on comparing the results (factual data) of written and video studies of dog training.

My definition of dog training is: To produce a dog that responds promptly, reliably, and willingly to verbal requests, when off-leash, distracted, at a distance, and without the continued need for any training tool, except for our voice.

In Stage 1, the RR%s largely reflect the dog's degree of comprehension of handsignals and verbal cues. Failure to fully understand the instruction in that scenario is by far the most common reason for non-compliance. While working through Stage 2, comprehension is pretty much a given, and the RR%s primarily reflect the dog's motivation to respond.

STAGE 2: To MOTIVATE Dogs to the Max — by primarily using considerably more powerful Life Rewards and the most effective reinforcement schedules — Random-Differential Reinforcement, i.e., to decrease the number of food rewards and thus, increase their reinforcing power

Reduce Food Rewards — by asking More for Less — more and more responses for fewer and fewer food rewards as monitored by Response:Reward Ratios, e.g., 7:1; and Longer for Less — longer and longer durations for fewer and fewer food rewards. (Count them out.) To give perspective, in Week One of puppy class, the average Golden will happily perform 20 puppy push-ups (Down-Sits) for the prospect of a single food reward, i.e., R:R Ratio = 40:1

Using LIFE REWARDS — by integrating very short training interludes, some as short as a single Sit, into your dog's favorite pastimes, such as playing, sniffing, and walks, and interactive games such as, Fetch, Tug, Tag, Find, Go To, Hide'n'Seek, etc. Every time you interrupt each activity/game, you may instruct your dog to resume. The more times you interrupt, the more times you may reinforce Sitting. So, ask for a Sit very 25 yards on walks, and every 30secs. or so during dog-dog play and interactive games.

For example, by frequently asking your dog to Sit during play, the sequence: 1. "Sit", 2. Take Collar, 3. Praise, 4. Take Tug or Kibble, followed by 5. "Go Play" becomes a cascade of reinforcers: 1. Quinary (Instruction to and Response of Sitting), 2. Quaternary (Collar grab — the #1 most common subliminal bite trigger), 3. Tertiary (PRAISE), 4. Secondary (Kibble or Tug*), followed by, 5. Primary Reinforcer (Instruct to Play and Playing). Such predictive, reinforcement cascades are very common in real life, for example, prior to feeding your dog, or taking them for a walk. (*I exclusively use a Tug toy amped up as a mega-secondary reinforcer when working with dog-dog reactivity, if I don't know how the other dog might react to food.)

We must Reinstate PRAISE! "G'dog" is both a precise 'marker' and a 'primary reinforcer', and one that may be amped up as an all-powerful, mega-secondary or tertiary reinforcer.
Moreover, praise is the only means of offering ongoing, differential reinforcement that mirrors ever-changing, long-duration behaviors, such as Stays, Watch, Off, and especially, PLAY. A single behavior never occurs in a vacuum and behavior is never static; behaviors run in packs, and behavior is in a state of constant flux, ever-changing rapidly, sometimes several times a second.

Having a mere quadrant of consequential feedback is too silly for words, considering that a dog could Sit in a zillion different ways. Our feedback must instantly reflect the degree and quality of every behavior change. The only way to accomplish this is by offering an ongoing running commentary. Certainly, your dog will not understand all your wonderful praise-adjectives, but they can read your affect, facial expressions, and body language like a book, when you speak the words.

STAGE 3: Resolve Misbehavior and Non-Compliance — by using our dog's vocabulary to redirect misbehavior to exactly what we would like the dog to do, and to curtail non-compliance via calm, loving, verbal negative reinforcement, a la kindergarten teachers and enlightened grandparents.

When dogs err (misbehavior and non-compliance), a single word conveys: 1. Stop what you're doing and 2. Do this instead, i.e., specifically what we would like the dog to do. For example, Sit, Settle, Shush, Steady. Chewtoy. Outside. All so simple; no need for distraction, DROs, or DRIs.

Verbal guidance is the Jewel in the Crown of lure-reward training and is a game-changer for dog professionals, dog owners, and their dogs. It must be so stressful for dogs when they have difficulty learning and see their owner's frustrations. To be guided by clear, comprehendible instructions and rewarded by happy, proud faces and praise from the heart must be such a relief for them. It is long overdue to embrace Bill Campbell's Jolly Routine, which was introduced 50-years before its time, and to this day, few people use it. It's a hoot! And dogs get it.

More Information about Lure-Reward Training
Please read my last dozen blogs at DunbarAcademy.com. You may download three books, and watch "Science-Based Dog Training (with feeling)" plus other courses — all for FREE by simply clicking https://dunbar.info/freepupbooks Also, check out the Spring 2024 issue of the APDT Chronicle of the Dog, or read my latest book, Barking Up the Right Tree. (I can't get you a freebie because that's one of my books that I didn't publish myself; but Amazon has decent prices.)

Also, starting in September, I have decided to offer a few seminars in the UK: dog-and-bone.co.uk and US dunbaracademy.com/pages/dr-dunbar-seminars about the various reward-based training techniques but focusing on lure-reward training, and dog social behavior (development and dog-dog reactivity). From the half a million comments on a recent post about a video showing me nipping puppy 'reactivity' in the bud, it was obvious that few people have a good grounding in normal puppy play behavior, especially, play-biting, play-fighting (chasing, lunging, growling, and scrapping) to be extremely worrisome. This is what puppies do within their litter, virtually all the time while awake: after a quick yawn, stretch, scratch, pee and maybe poop, they chase-chase-chase, bite-bite-bite and fight-fight-fight, until they suckle themselves to sleep. Here are brief descriptions while awake.

Peace in the Pack: The development of dog social behavior, especially the importance of puppy play-fighting when growing up... to reduce the need for fighting as adults and to eliminate the prospect of any serious injury (bite inhibition) should adult dogs have disagreements. Puppy-Adult interactions and Male-Female interactions are particularly fascinating especially, the 'Bitch Amendments to Male Hierarchical Law'. So intriguing, so true, and so effective.

Resolving Dog-Dog Reactivity: Very basic Lure-Reward Training, emphasizing the effectiveness of verbal feedback for guidance and reassurance (without unintentionally reinforcing your dog's outbursts). The secret for success is a couple of simple troubleshooting exercises to ensure your dog will respond reliably in any exciting or scary scenario. A simple Sit-Stay-Shush-Watch prevents your dog from lunging, barking and eyeballing the other dog. Your dog's waggy backside as it looks up at you is a much less threatening optic for other dogs and their owners; and feeling less threatened, they threaten your dog less.

Results-Based Dog Training: Again, primarily Lure-Reward Training but also, using other reward-based training techniques where appropriate, such as, Wait and Reward Training especially for out-of-control, over-the-top, hyperactive, rambunctious and rumbustious adolescents; Shaping for long-duration behaviors such as Stays, Watch, Shush, and Off;

Autoshaping for when the owner is away from home or otherwise preoccupied; and occasionally, Physical Prompting — using the oppositional reflex to solidify Sit-, Down-, and especially Stand-Stays.

Most trainers use only a single reward-based technique; I use all five. Sadly, the option of Electronic Autoshaping is no longer available, since, I suspect, my reward-only 'AutoTrainer' suffered a 'catch and kill'. A shame. Aside from teaching a dog to settle down and shush within 30 minutes, the machine nixed separation anxiety with just a couple of days.

Dog Training must re-invent itself.
 Please be a part of the revolution and help make the change.
 Let's get some real, practical science and fun and games back into playing The Training Game