Six Simple Steps of Positive Reinforcement Dog Training

Positive reinforcement dog training doesn’t have to be complex

Introduction

A major challenge in effective canine training is that it can seem to be overly complex, with debates about methodologies, procedures and tools, just to name some key variables. It often seems that the more you know, the more convoluted the suggested procedures appear to be. All of this commonly leaves would-be dog trainers wondering which way to turn, floundering in a sea of behavioural psychology, learning theories, scientific evidence, and practical advice – all hotly debated by numerous experts.

The simplified Six Step Model suggested here aims to cut through all of that and offer a baseline of simple and effective dog training in an easy to follow, graduated procedure, based on positive reinforcement principles[1].

Overview

The Six Step Model guides the trainer through the basics of teaching all of the fundamentals of what are often referred to as “dog manners” or “good citizen behaviour”. The same concepts can be extended to all forms of dog training, utilising the core theme of demonstrating the requirement or behaviour/action, adding a cue, marking the correct outcome and rewarding the dog. This procedure utilises “luring” as the starting point in the program –– and then extending the training goals to more complex actions/behaviours utilising “shaping” and “chaining” techniques”. Not all positive reinforcement training procedures use luring, for example some use “touch”, which is rather like “stationary luring”[2].  Coercive (and some “balanced”) trainers generally do not use luring, but focus more on such techniques as negative reinforcement and negative punishment (at the milder end of the spectrum) and positive punishment[3].

Dog learning can be thought of in four phases, which are: 1) Acquisition; 2) Fluency; 3) Generalisation; 4) Maintenance. Acquisition is the initial learning, associating the cue with the action and reward. Fluency is performing that desired action every time it is cued, not just some of the time. Generalisation is performing the action when cued under any circumstances, different environments and in the face of distractions. Maintenance is then reminding and maintaining the response to the cue over the long haul. The Six Steps described here obviously apply to the first three of these learning phases, but can also be applied to the fourth, reverting occasionally to the basic steps as a reminder (retraining) if required.

The six steps

The Six Step Model is shown graphically in Figure 1 below.

Like all models, this one is presented as a means of readily comprehending and applying a procedure which will deliver desired outcomes. Clearly, one adapts the model to fit each stage of the process, the dog’s response and the challenges which arise. This is no different than for any canine training procedure. The critical goal is implementation with a view to canine and human welfare as part of the end product, which involves starting with building and then maintaining the human-canine relationship, bond and trust and sequentially adding sustainable training outcomes.

At the start, the time spent on building the human-canine relationship, bond and trust is taken for granted (see companion Training Article on this website Zen and the art of companion dog training).

The details of implementation of the six sequential steps or stages of the training protocol are listed in order below.

1.     SHOW – LURE/HAND CUE: A small food treat is held in the flat or slightly cupped palm of the hand (may simply squeeze the treat in the palm or use the thumb to hold it in place) and placed just in front of the dog’s nose – then moved slowly away/around to the position desired to place the dog in. For example, with the dog in front of the trainer, the lure hand is shown to the dog then moved in a circular motion to the trainer’s left side (dog facing forward) then raised slightly above and over the back of the dog’s head – bingo! – the dog is now sitting at the “heel” position. The treat is not released until the end, that is, when the dog sits.

2.     SAY – VERBAL CUE: As the pup/dog shows that it understands this nice game of following the treat in the hand, this is the time to introduce the verbal cue the trainer wishes to associate with the action, such as verbalising “Heel” as the dog is lured to the trainer’s left side or “Sit” as the lure is raised slightly above and behind the dog’s head[4]. Note that the objective is to only deliver the verbal cue once as the dog comes to the desired position – the exercise will be repeated a number of times to enable the dog’s learning, but in each iteration the verbal cue (such as “Sit”) is only used once.

3.     SALARY – MARK AND PAY: The lure is used as described for Steps 1 and 2, then as the dog reaches the desired position (performs to desired behaviour) that moment is “marked” by either a “Click” using a training clicker or verbalised with a “Yes” (stated in a brief, upbeat voice). The timing here is quite important – as soon as the dog reaches the position/performs the action, or at least within about one second - that is the moment to “mark” that result, signifying to the dog that just at that moment it did exactly what you had cued/asked of it and now it will be rewarded, or paid, with a treat. This links with the “payment of salary” for performance, for which timing is also critical, as the “mark” (sometimes called the “bridge” or the action of “bridging” between the action and reward) is given then the payment comes within about 2 to 3 seconds of that “mark” (that is, the treat is released from the hand luring the dog after a very slight pause, not synchronously and not immediately).

4.     SIGNAL - HAND CUE: Now that the dog is following the hand and performing certain desired actions the trainer has cued, this is the time to guide the dog to the desired position with the hand – BUT – without the treat held in that hand. So now the simple hand action, say, for the sit, is the same as with the lure but is now just a vertical (and slightly back) hand motion. As the dog has learned this with the lure, when it “sits’ the trainer “marks” that action (“Click” or “Yes”) immediately then – within about 2 – 3 seconds – fishes a treat from their pouch/pocket and “pays” (rewards) the dog. At this stage the dog has learned two cues – one verbal and one hand gesture (body language).[5] The dog has also learned that the “Click” or “Yes” signifies that they did the right action and will be rewarded for it.

5.     SHAPE – SHAPE & CHAIN: “Shaping” in dog training – often called “shaping by progressive approximations” – involves achieving the training goal through small increments which build up to show the dog the final action/behaviour which is sought. For example, if the goal is to have the dog stand on a small box (upturned basin) with its front feet, the dog is progressively rewarded for looking at the box, touching/poking it with their nose, placing one foot then both feet on the box. Each iteration and progression is marked (“Click” or “Yes”) then rewarded – even though the final goal is yet to be achieved.[6] The method of “chaining” takes shaping a step further to build more complex actions which are progressively built up (in a chain of actions). For example, if the goal is to have the dog hold an object (such as a float bumper with an attached rope) in its mouth and carry it some distance to the box or upturned basin (or any other object/person), then the chaining procedure will involve teaching the ”hold the object”, then the next step, when that is achieved, is to teach the dog to carry the object progressive distances, then (separately) place its feet on the box/basin, then combine each of these “links in the chain” sequentially to create “pick up the object and run some distance to the box/basin and with front feet stand still on it and hold the object".

6.     SAFETY – WELFARE: The underpinning safety/welfare element of the training steps is a foundation element which is always front of mind. This involves an awareness that the training procedure must support canine and handler welfare, being cognisant of health, training environment, duration, frequency and is based on patience and persistence, all with a view to the challenge which the training goal presents.

The Six Steps Model described is further illustrated in Figure 2, showing the progressive nature of the procedure and how it may be duplicated for each new training goal.

 Conclusion

As in any dog training, consistency in delivery is required and the trainer needs to set their mindset to encompass realistic outcomes and expectations for their dog’s training. Always remember that the bonding step comes before anything else.

The advantages of this “Six Step Model” are that it is simple and easy to remember and logical in its progression, developing the dog’s learning in sequential steps. In addition, the model provides a plan which reduces time deciding on what’s next in the teaching program and eliminates frustration which can arise from moving too quickly in the training plan and “leaving the dog behind”.

The model does require a small amount of thought to consider the training outcomes and which goal to start with – but the advice is straight forward – start with the easy actions/behaviours (such as “follow-my-hand to the heel position” and “Sit”) and progress from there – don’t jump straight into chaining a complex manoeuvre – that will come from what you progressively build. Remembering the six progressive steps is not difficult, but the figures might be printed out as a “cheat sheet” to refer to when starting out.

 References

[1] With acknowledgement of the “6 Sigma Model” (Six Sigma TM USA) associated with the Japanese Toyota Production System (“TPS”) which was initiated as the 5S then became the 6S when Safety was added (paralleled here by “Welfare”). This is not suggesting that dog training parallels car manufacturing – the model is adapted because it is a proven, simple and successful way to break down a procedure into component steps which deliver the desired outcome. The original 6S system entails (with Japanese nomenclature): Seiri (sort), Seiton (set in order), Seiso (shine), Seiketsu (standardize), Shitsuke (sustain), and Safety (safety): e.g., URL https://www.leanproduction.com/5s/5s-history-and-manufacturing/

[2] For example, see the YouTube podcast between Susan Garret and Matt Folsom at https://youtu.be/327yASVbU4s?si=sQei1mawttI_Y5U2 

[3] See companion Article on this website The Operant Conditioning Model Fallacy for an explanation of these terms and the pros and cons of applying these techniques.

[4] Some trainers prefer to deliver the verbal cue first, before the lure action. The order of events presented here (physical luring gesture then later/sequentially adding the verbal cue) is based on the understanding that dogs learn and follow body language (i.e., gestures) more readily than learning words. We do find in practice that dogs learn to associate – and differentiate – the verbal and body language cues and with training will respond when used separately or together. Also see Footnote 5.

[5] Dogs are actually better at (quicker to read) body language than verbal cues, whereas for humans it is generally the opposite. The companion Article on this site discusses dog body language in Dog Body Language.  This presents a slight dilemma in terms of whether to introduce hand gestures/body language first and verbal cues second, or vice versa. The good news is that this is not supercritical, as in the end a trained dog will respond to either or both at once (humans tend to most readily fall to using the verbal and body gesture together as a preference, so the dog soon learns the association and responds to either when used singularly).

[6] Some trainers believe this method is preferable to luring the dog – for example luring it to put its front feet on the box – because it “creates a thinking dog”, whose actions guide its understanding of what the behaviour we want is: e.g., Garrett, S, (2018) Dog Training Blog: “Lurers Anonymous (Luring vs Shaping)”; URL https://susangarrettdogagility.com/2018/11/lurers-anonymous/. Either method will achieve the end result – luring guides the dog, whereas shaping requires the dog to progressively figure out what is being asked.