Clicker Training – An overview

by Elaine Clark IMDT

 

Why a clicker is a better “marker” than the voice

 

Quite simply, the clicker sound is always constant (your voice and the words you use change depending upon how you are feeling – stressed, tired, exasperated, happy ….). Also the click is much sharper and instantaneous. 

 

 

How Dogs “see” the clicker

 

The easiest way to explain this is to take the “Fruit Machine” as an analogy.  When you play a fruit machine you know you have to do something to have a chance of winning.  So, if we are training a trick, such as giving a paw …..

 

  1. You put in some money and press a button (your dog slightly lifts a paw).
  2. You look at the spinning fruit (dog waits to hear a click) and …..
  3. if you get 3 apples (click!) you know you have got it right and look for the money (reward).
  4. You put more money in and press the button again (your dog offers something else – e.glays down)
  5. You get 2 apples and a plum (no click) you don’t look for money, but try again (1 – 3).

 

The great bit for dogs and correct use of a clicker by the handler is that they soon work out that offering the same behaviour again is a sure-fired way of winning! If only “how” we pressed the button could ensure we won the jackpot again!!!!

 

The more clicker training a dog does the quicker she learns each new behaviour. Why? Because she has learnt that thinking brings results and that failure is only temporary and has gained the confidence to keep trying. She has also learnt that Mum or Dad is a great source of entertainment, is calm because she has used her brain and is, therefore, more likely to look at and listen to her parents in every environment. When I mention the word “training” my girls get more excited than when I pick up the leads (I don’t say “Walkies” for a very good reason, but that’s another story).


A popular misconception is that, by using a clicker and reward, your dog will learn only to respond to a cue if you have the clicker and reward in your hand. In fact, once a behaviour has been learnt (and we have proofed it) we put the rewards on an intermittent schedule and gradually fade them out altogether.

 

If this step is missed out, the dog sees the clicker as a vending machine, instead.  And we all know what would happen the first time you put money into that, press the button, look down and find nothing - you certainly would not have another go!!!!

How the trainer should see the clicker

 

The easiest way to explain this is to think of a camera.  If you want a photo of your dog mid-jump (in the air) ……

 

  1. click too early and your dog is just taking off,
  2. click too late and your dog has landed. 

 

If you want to take a photo of your dog sitting, rolling over, giving a paw etc. it is the same scenario. For “give a paw” the photo should show the paw not touching the floor – which means you have to be quick at first! (It doesn’t matter if the paw is back on the floor when the reward is delivered.)

 

Luring versus shaping

 

  1. Luring means you use a lure (food, toy – never force your dog into a position!) to get your dog to perform the required behaviour. You then give that reward. I may use this to begin teaching a “down”, or to begin teaching the agility tunnel or jump, for example. However, you need to very quickly phase out the lure or you are using bribes and you will always need to use them.
    There is a big difference between a bribe and a reward – the bribe comes before the behaviour as an incentive, whilst a reward comes after the behaviour which was freely offered. With a bribe the dog is thinking about the bribe and not learning how to earn a reward and you are in the vending machine scenario.
  2. Shaping naturally occurring behaviours – one way is to wait for your dog to naturally perform the targeted behaviour (you’re watching telly in the evening, clicker and treats ready, dog sits because he was born knowing how to sit, but not on cue …. Click and treat, but say nothing at first). With shaping we do not “name” a behaviour until it is consistently offered in response to a training environment. We ignore any other behaviours offered during that session.  So if you want your dog to shut a drawer with her nose and she offers you a “beg” or a “roll over” or goes and makes the bed and a cup of tea ….. Ignore! Lol ! 
  3. Free Shaping (increments of the behaviour we want) - we play the “Cold, luke warm, warmish, fairly hot, quite hot, hot and “WOOHOO!” game so the dog shouldn’t get it wrong. Before you begin, make a plan of all the tiny steps you can reward and make sure you are ready with clicker and treats. Keep sessions short. Keep to your training criteria for that session and if your dog “stalls out” rethink your set up.  If you click at the wrong time, reward anyway (your mistake – to not reward would devalue the click) and then be careful to get it right in future.

 

Either ask someone to observe or set up a camera to video the session – I don’t always see my mistakes, but I see my dogs making mistakes and so I know I have made one! Our dogs are much quicker to respond and much more consistent and accurate in the moment than we could ever hope to be.

 

 

Weaning off the clicker

 

Once your dog has thoroughly learnt a behaviour and the cue* you then need to begin the “weaning off clicker”. This is VERY important …. otherwise your dog will only work if you have a clicker and treats in view! (* can be verbal or a gesture.) So, let’s take “Sit!” as an example ….

 

  1. (This stage begins once she is reliably offering a sit in response to your cue).  Click and reward in position a few times. Encourage her to move out of the sit.
  2. Say “Sit!” then praise verbally and with a chuckle under the chin and encourage her to move.
  3. Repeat cue (verbal or gesture), click and treat. (So she has offered the behaviour twice before getting the click and treat).
  4. When clicking and treating you can also throw the treat a little distance on the floor so that the dog has to get up to collect their reward …. giving you the opportunity to ask for another sit, however in between clicks you will need to praise verbally and or physically and then encourage the dog to get up.
  5. Continue like that, but “ping pong” the number of “sits” (or other behaviour) required before the click and reward. By that I don’t mean keep on steadily-increasing the number of commands until you can ask for 20 sits before clicking and treating (because you will “lose” your dog’s attention waaaaaayyyyyy before you get there and then blame them and lose your temper and lose your dog’s trust and respect). I mean treat after 1, then 2, then 1, then 3, then 2, then 1, then 3, then 1, then 2, then 4, then 2, then 1, then 3 ………………. the most important is number one at first, but you can work towards 3 sits being the minimum, for example.
    BUT HUGE BUT – once the particular behaviour is thoroughly learnt you should occasionally reward it (pat, play, “Good girl” and food all work if mixed up often enough).

Because she will have had re-inforcement from you every now and then, later in the dog’s life you will only need to revisit each lesson occasionally with the clicker. (it becomes a habit ….. my dogs sit and wait at the side of the lane when a car passes, they sometimes get a food treat, sometimes a pat, sometimes “good girl” …. it is a habit for me and them.) 

  1. Result is that your dog will offer the “sit” (or other behaviour) quickly each time asked, just in case this time the reward (food, pat, verbal, game) will come -  keep playing the fruit machine!
    Paul and I don’t play the lottery anymore ….. guess why??!!! Because we only won a couple of tenners in many years of playing! So there need to be rewards at some level to keep a behaviour “alive”, but not too often as to cause extinction the first time it is inadvertently not rewarded – vending machine versus fruit machine!

A quick word about generalisation

 

People naturally generalise skills they have learnt – if you learn to speak French in the classroom, you can also speak French in the restaurant. This is not true for dogs when it comes to skills we have taught them (actually, some things they do generalise – things they have learnt themselves by association, fear for example).

So, for every new skill you teach your dog, take your sessions “out on tour” ..... practice at home, in different rooms, in the garden, at the park, in the woods, in town, at the vet – in fact in all the places you would normally need your dog to use that skill.  That extends not only to the environment (different locations, differing levels of distraction), but even to your own body position and your position in relation to your dog and whether they work one-to-one with you, with another family member or with other dogs they live with or walk with.

Have fun learning with your best friend!

 

Elaine Clark IMDT

Mon Copain mon Chien - The trainer your dog would choose.